Generative AI & LLM Engineer
For developers who want to build production LLM systems. You will be able to fine-tune models, build RAG pipelines and agents, evaluate and secure them, and serve them efficiently at scale.
{"nodes":[{"id":"title","type":"title","position":{"x":0,"y":0},"data":{"label":"Generative AI & LLM Engineer"},"width":1688,"height":61,"style":{"width":1688,"height":61},"measured":{"width":1688,"height":61},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":0}},{"id":"summary","type":"paragraph","position":{"x":404,"y":77},"data":{"label":"For developers who want to build production LLM systems. You will be able to fine-tune models, build RAG pipelines and agents, evaluate and secure them, and serve them efficiently at scale."},"width":880,"height":56,"style":{"width":880,"height":56},"measured":{"width":880,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":404,"y":77}},{"id":"meta","type":"paragraph","position":{"x":524,"y":147},"data":{"label":"10 STAGES · 35 TOPICS · 151 SUBTOPICS","style":{"fontSize":13,"fontFamily":"jetbrains","fontWeight":500,"color":"var(--color-fg-subtle)"}},"width":640,"height":21,"style":{"width":640,"height":21},"measured":{"width":640,"height":21},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":524,"y":147}},{"id":"stage-1","type":"section","position":{"x":0,"y":224},"data":{"label":"Prerequisites","description":"Build solid programming, math, and classical ML foundations for working with LLMs.","number":1},"width":824,"height":1016,"style":{"width":824,"height":1016},"measured":{"width":824,"height":1016},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":224}},{"id":"topic-1-1","type":"topic","position":{"x":28,"y":345},"data":{"label":"Programming Fundamentals","number":"1.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":345}},{"id":"sub-1-1-1","type":"subtopic","position":{"x":28,"y":413},"data":{"label":"Python Mastery","description":"Generators · Decorators · Typing"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":413}},{"id":"sub-1-1-2","type":"subtopic","position":{"x":288,"y":413},"data":{"label":"Concurrency & Async Programming","description":"asyncio · Threads · Processes"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":413}},{"id":"sub-1-1-3","type":"subtopic","position":{"x":548,"y":413},"data":{"label":"Object-Oriented Programming (OOP)"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":413}},{"id":"sub-1-1-4","type":"subtopic","position":{"x":28,"y":512},"data":{"label":"Data Structures & Algorithms"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":512}},{"id":"sub-1-1-5","type":"subtopic","position":{"x":288,"y":512},"data":{"label":"HTTP APIs","description":"REST · JSON · Streaming"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":512}},{"id":"sub-1-1-6","type":"subtopic","position":{"x":548,"y":512},"data":{"label":"Version Control","description":"Git · GitHub"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":512}},{"id":"sub-1-1-7","type":"subtopic","position":{"x":28,"y":591},"data":{"label":"Command Line & Bash Scripting"},"width":248,"height":65,"style":{"width":248,"height":65},"measured":{"width":248,"height":65},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":591}},{"id":"topic-1-2","type":"topic","position":{"x":28,"y":684},"data":{"label":"Mathematical Foundations","number":"1.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":684}},{"id":"sub-1-2-1","type":"subtopic","position":{"x":28,"y":752},"data":{"label":"Linear Algebra","description":"Vectors · Matrices · Tensors"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":752}},{"id":"sub-1-2-2","type":"subtopic","position":{"x":288,"y":752},"data":{"label":"Multivariable Calculus","description":"Derivatives · Gradients · Chain Rule"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":752}},{"id":"sub-1-2-3","type":"subtopic","position":{"x":548,"y":752},"data":{"label":"Probability & Statistics","description":"Distributions · Bayes' Theorem · Sampling"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":752}},{"id":"sub-1-2-4","type":"subtopic","position":{"x":28,"y":849},"data":{"label":"Information Theory","description":"Entropy · Cross-Entropy · KL Divergence"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":849}},{"id":"sub-1-2-5","type":"subtopic","position":{"x":288,"y":849},"data":{"label":"Optimization Basics","description":"Gradient Descent · Convexity"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":849}},{"id":"topic-1-3","type":"topic","position":{"x":28,"y":962},"data":{"label":"Traditional Machine Learning","number":"1.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":962}},{"id":"sub-1-3-1","type":"subtopic","position":{"x":28,"y":1030},"data":{"label":"Supervised vs Unsupervised Learning"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1030}},{"id":"sub-1-3-2","type":"subtopic","position":{"x":288,"y":1030},"data":{"label":"scikit-learn Framework"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":1030}},{"id":"sub-1-3-3","type":"subtopic","position":{"x":548,"y":1030},"data":{"label":"Evaluation Metrics","description":"Accuracy · Precision · Recall · F1"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":1030}},{"id":"sub-1-3-4","type":"subtopic","position":{"x":28,"y":1127},"data":{"label":"Overfitting & Regularization","description":"Bias-Variance · L1 · L2 · Dropout"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1127}},{"id":"stage-2","type":"section","position":{"x":864,"y":224},"data":{"label":"Deep Learning & NLP Fundamentals","description":"Learn neural network training and the classic NLP ideas that led to Transformers.","number":2},"width":824,"height":1016,"style":{"width":824,"height":1016},"measured":{"width":824,"height":1016},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":864,"y":224}},{"id":"topic-2-1","type":"topic","position":{"x":892,"y":345},"data":{"label":"Deep Learning Basics","number":"2.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":345}},{"id":"sub-2-1-1","type":"subtopic","position":{"x":892,"y":413},"data":{"label":"Artificial Neural Networks (ANNs)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":413}},{"id":"sub-2-1-2","type":"subtopic","position":{"x":1152,"y":413},"data":{"label":"Backpropagation & Loss Functions"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":413}},{"id":"sub-2-1-3","type":"subtopic","position":{"x":1412,"y":413},"data":{"label":"Activation Functions","description":"ReLU · GELU · SiLU"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":413}},{"id":"sub-2-1-4","type":"subtopic","position":{"x":892,"y":492},"data":{"label":"Optimizers & Schedulers","description":"AdamW · Warmup · Cosine Annealing"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":492}},{"id":"sub-2-1-5","type":"subtopic","position":{"x":1152,"y":492},"data":{"label":"Deep Learning Frameworks","description":"PyTorch · JAX"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":492}},{"id":"topic-2-2","type":"topic","position":{"x":892,"y":605},"data":{"label":"Classical NLP","number":"2.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":605}},{"id":"sub-2-2-1","type":"subtopic","position":{"x":892,"y":673},"data":{"label":"Text Preprocessing","description":"Word Tokenization · Stemming · Lemmatization"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":673}},{"id":"sub-2-2-2","type":"subtopic","position":{"x":1152,"y":673},"data":{"label":"Bag of Words & TF-IDF"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":673}},{"id":"sub-2-2-3","type":"subtopic","position":{"x":1412,"y":673},"data":{"label":"Word Embeddings","description":"Word2Vec · GloVe · FastText"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":673}},{"id":"sub-2-2-4","type":"subtopic","position":{"x":892,"y":770},"data":{"label":"Subword Tokenization","description":"BPE · WordPiece · SentencePiece · tiktoken"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":770}},{"id":"topic-2-3","type":"topic","position":{"x":892,"y":883},"data":{"label":"Sequence Models","number":"2.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":883}},{"id":"sub-2-3-1","type":"subtopic","position":{"x":892,"y":951},"data":{"label":"Recurrent Neural Networks (RNNs) & LSTMs"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":951}},{"id":"sub-2-3-2","type":"subtopic","position":{"x":1152,"y":951},"data":{"label":"Encoder-Decoder Architecture (Seq2Seq)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":951}},{"id":"sub-2-3-3","type":"subtopic","position":{"x":1412,"y":951},"data":{"label":"Attention Mechanism","description":"Additive · Multiplicative"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":951}},{"id":"stage-3","type":"section","position":{"x":0,"y":1280},"data":{"label":"The Transformer Architecture","description":"Understand every Transformer component, its main variants, and the tooling to use them.","number":3},"width":824,"height":982,"style":{"width":824,"height":982},"measured":{"width":824,"height":982},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":1280}},{"id":"topic-3-1","type":"topic","position":{"x":28,"y":1401},"data":{"label":"Transformer Anatomy","number":"3.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1401}},{"id":"sub-3-1-1","type":"subtopic","position":{"x":28,"y":1469},"data":{"label":"Self-Attention & Multi-Head Attention"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1469}},{"id":"sub-3-1-2","type":"subtopic","position":{"x":288,"y":1469},"data":{"label":"Positional Encoding","description":"Absolute · Relative · RoPE · ALiBi"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":1469}},{"id":"sub-3-1-3","type":"subtopic","position":{"x":548,"y":1469},"data":{"label":"Layer Normalization","description":"Post-Norm · Pre-Norm · RMSNorm"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":1469}},{"id":"sub-3-1-4","type":"subtopic","position":{"x":28,"y":1566},"data":{"label":"Feed-Forward Layers","description":"MLP · SwiGLU"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1566}},{"id":"sub-3-1-5","type":"subtopic","position":{"x":288,"y":1566},"data":{"label":"Efficient Attention","description":"MQA · GQA · MLA · FlashAttention"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":1566}},{"id":"topic-3-2","type":"topic","position":{"x":28,"y":1679},"data":{"label":"Transformer Variants","number":"3.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1679}},{"id":"sub-3-2-1","type":"subtopic","position":{"x":28,"y":1747},"data":{"label":"Encoder-Only Models","description":"BERT · RoBERTa · ModernBERT"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1747}},{"id":"sub-3-2-2","type":"subtopic","position":{"x":288,"y":1747},"data":{"label":"Decoder-Only Models","description":"GPT · Llama · Mistral · Qwen · DeepSeek"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":1747}},{"id":"sub-3-2-3","type":"subtopic","position":{"x":548,"y":1747},"data":{"label":"Encoder-Decoder Models","description":"T5 · BART"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":1747}},{"id":"sub-3-2-4","type":"subtopic","position":{"x":28,"y":1844},"data":{"label":"Mixture of Experts (MoE)"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1844}},{"id":"topic-3-3","type":"topic","position":{"x":28,"y":1918},"data":{"label":"Alternative Architectures","number":"3.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1918}},{"id":"sub-3-3-1","type":"subtopic","position":{"x":28,"y":1986},"data":{"label":"State Space Models","description":"Mamba · Mamba-2"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":1986}},{"id":"sub-3-3-2","type":"subtopic","position":{"x":288,"y":1986},"data":{"label":"Hybrid Architectures (Jamba)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":1986}},{"id":"sub-3-3-3","type":"subtopic","position":{"x":548,"y":1986},"data":{"label":"RWKV & Linear Attention"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":1986}},{"id":"topic-3-4","type":"topic","position":{"x":28,"y":2081},"data":{"label":"Hugging Face Ecosystem","number":"3.4"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2081}},{"id":"sub-3-4-1","type":"subtopic","position":{"x":28,"y":2149},"data":{"label":"Transformers Library","description":"Pipelines · AutoModel · Tokenizers"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2149}},{"id":"sub-3-4-2","type":"subtopic","position":{"x":288,"y":2149},"data":{"label":"Datasets Library"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":2149}},{"id":"sub-3-4-3","type":"subtopic","position":{"x":548,"y":2149},"data":{"label":"Model Hubs","description":"Hugging Face Hub · ModelScope"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":2149}},{"id":"stage-4","type":"section","position":{"x":864,"y":1280},"data":{"label":"LLM Training & Fine-Tuning","description":"Learn how LLMs are pretrained, aligned, and efficiently adapted to new tasks.","number":4},"width":824,"height":982,"style":{"width":824,"height":982},"measured":{"width":824,"height":982},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":864,"y":1280}},{"id":"topic-4-1","type":"topic","position":{"x":892,"y":1401},"data":{"label":"Pre-training","number":"4.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1401}},{"id":"sub-4-1-1","type":"subtopic","position":{"x":892,"y":1469},"data":{"label":"Pre-training Objectives","description":"Next-Token Prediction · Masked LM"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1469}},{"id":"sub-4-1-2","type":"subtopic","position":{"x":1152,"y":1469},"data":{"label":"Data Curation","description":"Filtering · Deduplication · PII Scrubbing"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":1469}},{"id":"sub-4-1-3","type":"subtopic","position":{"x":1412,"y":1469},"data":{"label":"Scaling Laws","description":"Chinchilla · Compute-Optimal Training"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":1469}},{"id":"sub-4-1-4","type":"subtopic","position":{"x":892,"y":1566},"data":{"label":"Distributed Training","description":"Data · Tensor · Pipeline Parallelism · FSDP · DeepSpeed"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1566}},{"id":"topic-4-2","type":"topic","position":{"x":892,"y":1679},"data":{"label":"Post-Training & Alignment","number":"4.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1679}},{"id":"sub-4-2-1","type":"subtopic","position":{"x":892,"y":1747},"data":{"label":"Instruction Tuning","description":"SFT · Chat Templates"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1747}},{"id":"sub-4-2-2","type":"subtopic","position":{"x":1152,"y":1747},"data":{"label":"Synthetic Data","description":"Self-Instruct · Evol-Instruct · Distillation"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":1747}},{"id":"sub-4-2-3","type":"subtopic","position":{"x":1412,"y":1747},"data":{"label":"RLHF","description":"Reward Modeling · PPO"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":1747}},{"id":"sub-4-2-4","type":"subtopic","position":{"x":892,"y":1844},"data":{"label":"Preference Optimization","description":"DPO · ORPO · KTO · SimPO"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1844}},{"id":"sub-4-2-5","type":"subtopic","position":{"x":1152,"y":1844},"data":{"label":"Constitutional AI & RLAIF"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":1844}},{"id":"sub-4-2-6","type":"subtopic","position":{"x":1412,"y":1844},"data":{"label":"Reasoning Models","description":"GRPO · Verifiable Rewards · Test-Time Compute"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":1844}},{"id":"topic-4-3","type":"topic","position":{"x":892,"y":1957},"data":{"label":"Parameter-Efficient Fine-Tuning (PEFT)","number":"4.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":1957}},{"id":"sub-4-3-1","type":"subtopic","position":{"x":892,"y":2025},"data":{"label":"Low-Rank Adaptation (LoRA)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2025}},{"id":"sub-4-3-2","type":"subtopic","position":{"x":1152,"y":2025},"data":{"label":"LoRA Variants","description":"QLoRA · DoRA · PiSSA"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2025}},{"id":"sub-4-3-3","type":"subtopic","position":{"x":1412,"y":2025},"data":{"label":"Prompt Tuning & Prefix Tuning"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":2025}},{"id":"sub-4-3-4","type":"subtopic","position":{"x":892,"y":2104},"data":{"label":"Adapter Layers"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2104}},{"id":"sub-4-3-5","type":"subtopic","position":{"x":1152,"y":2104},"data":{"label":"Fine-Tuning Tools","description":"PEFT · TRL · Unsloth · Axolotl"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2104}},{"id":"stage-5","type":"section","position":{"x":0,"y":2301},"data":{"label":"Building LLM Applications","description":"Prompt models effectively and ground them in your data with robust RAG pipelines.","number":5},"width":824,"height":1218,"style":{"width":824,"height":1218},"measured":{"width":824,"height":1218},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":2301}},{"id":"topic-5-1","type":"topic","position":{"x":28,"y":2422},"data":{"label":"Prompt Engineering","number":"5.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2422}},{"id":"sub-5-1-1","type":"subtopic","position":{"x":28,"y":2490},"data":{"label":"Zero-Shot & Few-Shot Prompting"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2490}},{"id":"sub-5-1-2","type":"subtopic","position":{"x":288,"y":2490},"data":{"label":"Chain-of-Thought (CoT)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":2490}},{"id":"sub-5-1-3","type":"subtopic","position":{"x":548,"y":2490},"data":{"label":"Tree & Graph of Thoughts","description":"ToT · GoT"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":2490}},{"id":"sub-5-1-4","type":"subtopic","position":{"x":28,"y":2569},"data":{"label":"System Prompts & Personas"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2569}},{"id":"sub-5-1-5","type":"subtopic","position":{"x":288,"y":2569},"data":{"label":"Structured Outputs","description":"JSON Mode · JSON Schema · Pydantic"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":2569}},{"id":"sub-5-1-6","type":"subtopic","position":{"x":548,"y":2569},"data":{"label":"LLM Provider APIs","description":"OpenAI · Anthropic · Gemini"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":2569}},{"id":"topic-5-2","type":"topic","position":{"x":28,"y":2682},"data":{"label":"Frameworks & Orchestration","number":"5.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2682}},{"id":"sub-5-2-1","type":"subtopic","position":{"x":28,"y":2750},"data":{"label":"LangChain","description":"Runnables · Prompts · Output Parsers"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2750}},{"id":"sub-5-2-2","type":"subtopic","position":{"x":288,"y":2750},"data":{"label":"LlamaIndex","description":"Data Connectors · Indices · Query Engines"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":2750}},{"id":"sub-5-2-3","type":"subtopic","position":{"x":548,"y":2750},"data":{"label":"Haystack","description":"Pipelines · Components"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":2750}},{"id":"sub-5-2-4","type":"subtopic","position":{"x":28,"y":2847},"data":{"label":"DSPy","description":"Signatures · Modules · Optimizers"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2847}},{"id":"topic-5-3","type":"topic","position":{"x":28,"y":2960},"data":{"label":"RAG Fundamentals","number":"5.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":2960}},{"id":"sub-5-3-1","type":"subtopic","position":{"x":28,"y":3028},"data":{"label":"Naive vs Advanced RAG"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3028}},{"id":"sub-5-3-2","type":"subtopic","position":{"x":288,"y":3028},"data":{"label":"Document Loading & Parsing","description":"PDF · HTML · Docling · Unstructured"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3028}},{"id":"sub-5-3-3","type":"subtopic","position":{"x":548,"y":3028},"data":{"label":"Chunking Strategies","description":"Fixed-Size · Recursive · Semantic · Agentic"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":3028}},{"id":"sub-5-3-4","type":"subtopic","position":{"x":28,"y":3146},"data":{"label":"Embedding Models","description":"OpenAI · BGE · Nomic · Voyage · Cohere"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3146}},{"id":"sub-5-3-5","type":"subtopic","position":{"x":288,"y":3146},"data":{"label":"Vector Databases","description":"Pinecone · Milvus · Qdrant · Chroma · pgvector"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3146}},{"id":"sub-5-3-6","type":"subtopic","position":{"x":548,"y":3146},"data":{"label":"Retrieval Strategies","description":"BM25 · Dense Retrieval · Hybrid Search"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":3146}},{"id":"topic-5-4","type":"topic","position":{"x":28,"y":3259},"data":{"label":"Advanced RAG","number":"5.4"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3259}},{"id":"sub-5-4-1","type":"subtopic","position":{"x":28,"y":3327},"data":{"label":"Query Transformation","description":"HyDE · Multi-Query · Semantic Routing"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3327}},{"id":"sub-5-4-2","type":"subtopic","position":{"x":288,"y":3327},"data":{"label":"Reranking","description":"Cross-Encoders · Cohere Rerank"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3327}},{"id":"sub-5-4-3","type":"subtopic","position":{"x":548,"y":3327},"data":{"label":"GraphRAG & Knowledge Graphs"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":3327}},{"id":"sub-5-4-4","type":"subtopic","position":{"x":28,"y":3424},"data":{"label":"Adaptive RAG","description":"Self-RAG · CRAG · Agentic RAG"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3424}},{"id":"sub-5-4-5","type":"subtopic","position":{"x":288,"y":3424},"data":{"label":"Context Compression (LLMLingua)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3424}},{"id":"stage-6","type":"section","position":{"x":864,"y":2301},"data":{"label":"AI Agents & Evaluation","description":"Build tool-using agents and measure LLM system quality with evals and tracing.","number":6},"width":824,"height":1218,"style":{"width":824,"height":1218},"measured":{"width":824,"height":1218},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":864,"y":2301}},{"id":"topic-6-1","type":"topic","position":{"x":892,"y":2422},"data":{"label":"Agent Fundamentals","number":"6.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2422}},{"id":"sub-6-1-1","type":"subtopic","position":{"x":892,"y":2490},"data":{"label":"ReAct Pattern","description":"Reason · Act · Observe"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2490}},{"id":"sub-6-1-2","type":"subtopic","position":{"x":1152,"y":2490},"data":{"label":"Function Calling & Tool Use"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2490}},{"id":"sub-6-1-3","type":"subtopic","position":{"x":1412,"y":2490},"data":{"label":"Model Context Protocol (MCP)"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":2490}},{"id":"sub-6-1-4","type":"subtopic","position":{"x":892,"y":2569},"data":{"label":"Plan-and-Execute Architectures"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2569}},{"id":"sub-6-1-5","type":"subtopic","position":{"x":1152,"y":2569},"data":{"label":"Agent Memory","description":"Short-Term · Long-Term · Entity"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2569}},{"id":"topic-6-2","type":"topic","position":{"x":892,"y":2664},"data":{"label":"Multi-Agent Systems","number":"6.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2664}},{"id":"sub-6-2-1","type":"subtopic","position":{"x":892,"y":2732},"data":{"label":"Agent Frameworks","description":"LangGraph · CrewAI · OpenAI Agents SDK · Microsoft Agent Framework"},"width":248,"height":104,"style":{"width":248,"height":104},"measured":{"width":248,"height":104},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2732}},{"id":"sub-6-2-2","type":"subtopic","position":{"x":1152,"y":2732},"data":{"label":"Multi-Agent Patterns","description":"Supervisor · Hierarchical · Handoffs"},"width":248,"height":104,"style":{"width":248,"height":104},"measured":{"width":248,"height":104},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2732}},{"id":"sub-6-2-3","type":"subtopic","position":{"x":1412,"y":2732},"data":{"label":"Execution Sandboxes","description":"Code Interpreters · Containers"},"width":248,"height":104,"style":{"width":248,"height":104},"measured":{"width":248,"height":104},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":2732}},{"id":"sub-6-2-4","type":"subtopic","position":{"x":892,"y":2848},"data":{"label":"Human-in-the-Loop"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2848}},{"id":"topic-6-3","type":"topic","position":{"x":892,"y":2922},"data":{"label":"Evaluation","number":"6.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2922}},{"id":"sub-6-3-1","type":"subtopic","position":{"x":892,"y":2990},"data":{"label":"Benchmarks & Leaderboards","description":"MMLU · HumanEval · MT-Bench · LMArena"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":2990}},{"id":"sub-6-3-2","type":"subtopic","position":{"x":1152,"y":2990},"data":{"label":"Text Metrics","description":"BLEU · ROUGE · BERTScore"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":2990}},{"id":"sub-6-3-3","type":"subtopic","position":{"x":1412,"y":2990},"data":{"label":"LLM-as-a-Judge","description":"G-Eval · Prometheus · Pairwise Grading"},"width":248,"height":106,"style":{"width":248,"height":106},"measured":{"width":248,"height":106},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":2990}},{"id":"sub-6-3-4","type":"subtopic","position":{"x":892,"y":3108},"data":{"label":"RAG Evaluation","description":"RAGAS · TruLens · DeepEval"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3108}},{"id":"sub-6-3-5","type":"subtopic","position":{"x":1152,"y":3108},"data":{"label":"Agent Evaluation","description":"Trajectories · Tool-Call Accuracy"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":3108}},{"id":"topic-6-4","type":"topic","position":{"x":892,"y":3221},"data":{"label":"Observability","number":"6.4"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3221}},{"id":"sub-6-4-1","type":"subtopic","position":{"x":892,"y":3289},"data":{"label":"Tracing","description":"LangSmith · Langfuse · Phoenix"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3289}},{"id":"sub-6-4-2","type":"subtopic","position":{"x":1152,"y":3289},"data":{"label":"Cost & Latency Monitoring","description":"Tokens · TTFT · Throughput"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":3289}},{"id":"sub-6-4-3","type":"subtopic","position":{"x":1412,"y":3289},"data":{"label":"A/B Testing & Shadow Deployments"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":3289}},{"id":"stage-7","type":"section","position":{"x":0,"y":3559},"data":{"label":"Security, Safety & Governance","description":"Defend LLM systems against attacks and ship them responsibly and compliantly.","number":7},"width":824,"height":1118,"style":{"width":824,"height":1118},"measured":{"width":824,"height":1118},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":3559}},{"id":"topic-7-1","type":"topic","position":{"x":28,"y":3680},"data":{"label":"Model Safety & Ethics","number":"7.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3680}},{"id":"sub-7-1-1","type":"subtopic","position":{"x":28,"y":3748},"data":{"label":"Bias, Fairness & Toxicity"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3748}},{"id":"sub-7-1-2","type":"subtopic","position":{"x":288,"y":3748},"data":{"label":"Hallucinations & Grounding"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3748}},{"id":"sub-7-1-3","type":"subtopic","position":{"x":548,"y":3748},"data":{"label":"Red Teaming","description":"Garak · PyRIT · promptfoo"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":3748}},{"id":"sub-7-1-4","type":"subtopic","position":{"x":28,"y":3827},"data":{"label":"Machine Unlearning & Copyright"},"width":248,"height":65,"style":{"width":248,"height":65},"measured":{"width":248,"height":65},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3827}},{"id":"topic-7-2","type":"topic","position":{"x":28,"y":3920},"data":{"label":"Attacks & Defenses","number":"7.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3920}},{"id":"sub-7-2-1","type":"subtopic","position":{"x":28,"y":3988},"data":{"label":"Prompt Injection","description":"Direct · Indirect"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":3988}},{"id":"sub-7-2-2","type":"subtopic","position":{"x":288,"y":3988},"data":{"label":"Jailbreaking"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":3988}},{"id":"sub-7-2-3","type":"subtopic","position":{"x":548,"y":3988},"data":{"label":"Data Poisoning"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":3988}},{"id":"sub-7-2-4","type":"subtopic","position":{"x":28,"y":4067},"data":{"label":"Input Validation & Defensive Prompting"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":4067}},{"id":"sub-7-2-5","type":"subtopic","position":{"x":288,"y":4067},"data":{"label":"Guardrails","description":"NeMo Guardrails · Llama Guard · Guardrails AI"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":4067}},{"id":"sub-7-2-6","type":"subtopic","position":{"x":548,"y":4067},"data":{"label":"Agent Security","description":"Least Privilege · Tool Permissions"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":4067}},{"id":"topic-7-3","type":"topic","position":{"x":28,"y":4180},"data":{"label":"Governance & Compliance","number":"7.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":4180}},{"id":"sub-7-3-1","type":"subtopic","position":{"x":28,"y":4248},"data":{"label":"Enterprise Compliance","description":"SOC 2 · GDPR · HIPAA"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":4248}},{"id":"sub-7-3-2","type":"subtopic","position":{"x":288,"y":4248},"data":{"label":"AI Regulation","description":"EU AI Act · NIST AI RMF"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":4248}},{"id":"sub-7-3-3","type":"subtopic","position":{"x":548,"y":4248},"data":{"label":"Documentation","description":"Model Cards · Datasheets"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":4248}},{"id":"stage-8","type":"section","position":{"x":864,"y":3559},"data":{"label":"Deployment, Inference & Optimization","description":"Serve LLMs fast and cost-effectively on GPUs, the cloud, and local machines.","number":8},"width":824,"height":1118,"style":{"width":824,"height":1118},"measured":{"width":824,"height":1118},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":864,"y":3559}},{"id":"topic-8-1","type":"topic","position":{"x":892,"y":3680},"data":{"label":"GPU & Compute Fundamentals","number":"8.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3680}},{"id":"sub-8-1-1","type":"subtopic","position":{"x":892,"y":3748},"data":{"label":"GPU Architecture","description":"CUDA Cores · Tensor Cores · HBM"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3748}},{"id":"sub-8-1-2","type":"subtopic","position":{"x":1152,"y":3748},"data":{"label":"Memory Hierarchy & Bandwidth"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":3748}},{"id":"sub-8-1-3","type":"subtopic","position":{"x":1412,"y":3748},"data":{"label":"Compute-Bound vs Memory-Bound Workloads"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":3748}},{"id":"sub-8-1-4","type":"subtopic","position":{"x":892,"y":3845},"data":{"label":"VRAM Estimation","description":"Weights · KV Cache · Activations"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3845}},{"id":"topic-8-2","type":"topic","position":{"x":892,"y":3958},"data":{"label":"Optimization Techniques","number":"8.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":3958}},{"id":"sub-8-2-1","type":"subtopic","position":{"x":892,"y":4026},"data":{"label":"Quantization","description":"GPTQ · AWQ · GGUF · FP8 · SmoothQuant"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4026}},{"id":"sub-8-2-2","type":"subtopic","position":{"x":1152,"y":4026},"data":{"label":"Continuous Batching & PagedAttention"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4026}},{"id":"sub-8-2-3","type":"subtopic","position":{"x":1412,"y":4026},"data":{"label":"KV Cache Optimization & Prompt Caching"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":4026}},{"id":"sub-8-2-4","type":"subtopic","position":{"x":892,"y":4123},"data":{"label":"Speculative Decoding","description":"Draft Models · Medusa · EAGLE"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4123}},{"id":"sub-8-2-5","type":"subtopic","position":{"x":1152,"y":4123},"data":{"label":"Context Length Extension","description":"YaRN · RoPE Scaling"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4123}},{"id":"sub-8-2-6","type":"subtopic","position":{"x":1412,"y":4123},"data":{"label":"Pruning & Distillation"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":4123}},{"id":"topic-8-3","type":"topic","position":{"x":892,"y":4218},"data":{"label":"Serving Engines","number":"8.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4218}},{"id":"sub-8-3-1","type":"subtopic","position":{"x":892,"y":4286},"data":{"label":"High-Throughput Serving","description":"vLLM · SGLang"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4286}},{"id":"sub-8-3-2","type":"subtopic","position":{"x":1152,"y":4286},"data":{"label":"NVIDIA Stack","description":"TensorRT-LLM · Triton · NIM"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4286}},{"id":"sub-8-3-3","type":"subtopic","position":{"x":1412,"y":4286},"data":{"label":"Local Inference","description":"Ollama · llama.cpp · MLX · LM Studio"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":4286}},{"id":"topic-8-4","type":"topic","position":{"x":892,"y":4399},"data":{"label":"Cloud & Production Deployment","number":"8.4"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4399}},{"id":"sub-8-4-1","type":"subtopic","position":{"x":892,"y":4467},"data":{"label":"Managed Platforms","description":"AWS Bedrock · Azure AI Foundry · Vertex AI"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4467}},{"id":"sub-8-4-2","type":"subtopic","position":{"x":1152,"y":4467},"data":{"label":"Serverless GPUs","description":"RunPod · Modal · Replicate · Baseten"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4467}},{"id":"sub-8-4-3","type":"subtopic","position":{"x":1412,"y":4467},"data":{"label":"API Design","description":"Streaming · Rate Limiting · Caching"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":4467}},{"id":"sub-8-4-4","type":"subtopic","position":{"x":892,"y":4564},"data":{"label":"LLM Gateways","description":"LiteLLM · Model Routing · Fallbacks"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4564}},{"id":"stage-9","type":"section","position":{"x":0,"y":4716},"data":{"label":"Multimodal AI & Emerging Trends","description":"Extend LLM skills to vision, speech, and media generation, and track where the field is heading.","number":9},"width":824,"height":848,"style":{"width":824,"height":848},"measured":{"width":824,"height":848},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":0,"y":4716}},{"id":"topic-9-1","type":"topic","position":{"x":28,"y":4837},"data":{"label":"Vision-Language Models","number":"9.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":4837}},{"id":"sub-9-1-1","type":"subtopic","position":{"x":28,"y":4905},"data":{"label":"Contrastive Image-Text Models","description":"CLIP · SigLIP"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":4905}},{"id":"sub-9-1-2","type":"subtopic","position":{"x":288,"y":4905},"data":{"label":"Multimodal LLMs","description":"LLaVA · Qwen-VL · InternVL"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":4905}},{"id":"sub-9-1-3","type":"subtopic","position":{"x":548,"y":4905},"data":{"label":"Document AI","description":"LayoutLM · OCR-Free Parsing"},"width":248,"height":87,"style":{"width":248,"height":87},"measured":{"width":248,"height":87},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":4905}},{"id":"topic-9-2","type":"topic","position":{"x":28,"y":5020},"data":{"label":"Speech & Audio","number":"9.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5020}},{"id":"sub-9-2-1","type":"subtopic","position":{"x":28,"y":5088},"data":{"label":"Speech-to-Text (Whisper)"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5088}},{"id":"sub-9-2-2","type":"subtopic","position":{"x":288,"y":5088},"data":{"label":"Text-to-Speech (TTS)"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":5088}},{"id":"sub-9-2-3","type":"subtopic","position":{"x":548,"y":5088},"data":{"label":"Realtime Voice Agents"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":5088}},{"id":"topic-9-3","type":"topic","position":{"x":28,"y":5162},"data":{"label":"Image & Video Generation","number":"9.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5162}},{"id":"sub-9-3-1","type":"subtopic","position":{"x":28,"y":5230},"data":{"label":"Image Generation","description":"Stable Diffusion · Flux · Midjourney"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5230}},{"id":"sub-9-3-2","type":"subtopic","position":{"x":288,"y":5230},"data":{"label":"Video Generation","description":"Sora · Veo"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":5230}},{"id":"topic-9-4","type":"topic","position":{"x":28,"y":5343},"data":{"label":"Emerging Trends","number":"9.4"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5343}},{"id":"sub-9-4-1","type":"subtopic","position":{"x":28,"y":5411},"data":{"label":"Small Language Models & Edge AI"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5411}},{"id":"sub-9-4-2","type":"subtopic","position":{"x":288,"y":5411},"data":{"label":"Continual Learning in LLMs"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":288,"y":5411}},{"id":"sub-9-4-3","type":"subtopic","position":{"x":548,"y":5411},"data":{"label":"Autonomous Coding Agents","description":"Claude Code · Codex · Devin"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":548,"y":5411}},{"id":"sub-9-4-4","type":"subtopic","position":{"x":28,"y":5490},"data":{"label":"Computer-Use Agents"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":28,"y":5490}},{"id":"stage-10","type":"section","position":{"x":864,"y":4716},"data":{"label":"Capstone & Career","description":"Ship portfolio-grade LLM projects and prepare for GenAI engineering interviews.","number":10},"width":824,"height":848,"style":{"width":824,"height":848},"measured":{"width":824,"height":848},"zIndex":-999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":864,"y":4716}},{"id":"topic-10-1","type":"topic","position":{"x":892,"y":4837},"data":{"label":"Portfolio Projects","number":"10.1"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4837}},{"id":"sub-10-1-1","type":"subtopic","position":{"x":892,"y":4905},"data":{"label":"Production RAG Assistant"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4905}},{"id":"sub-10-1-2","type":"subtopic","position":{"x":1152,"y":4905},"data":{"label":"Multi-Agent Workflow"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4905}},{"id":"sub-10-1-3","type":"subtopic","position":{"x":1412,"y":4905},"data":{"label":"Fine-Tuned Domain Model","description":"LoRA · Evaluation · Deployment"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":4905}},{"id":"sub-10-1-4","type":"subtopic","position":{"x":892,"y":4984},"data":{"label":"LLM Evaluation Harness"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":4984}},{"id":"sub-10-1-5","type":"subtopic","position":{"x":1152,"y":4984},"data":{"label":"Open Source Contributions"},"width":248,"height":46,"style":{"width":248,"height":46},"measured":{"width":248,"height":46},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":4984}},{"id":"topic-10-2","type":"topic","position":{"x":892,"y":5058},"data":{"label":"Job Preparation","number":"10.2"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":5058}},{"id":"sub-10-2-1","type":"subtopic","position":{"x":892,"y":5126},"data":{"label":"Resume & Portfolio","description":"GitHub · Demos · Write-Ups"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":5126}},{"id":"sub-10-2-2","type":"subtopic","position":{"x":1152,"y":5126},"data":{"label":"LLM System Design Interviews"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":5126}},{"id":"sub-10-2-3","type":"subtopic","position":{"x":1412,"y":5126},"data":{"label":"Transformer & LLM Theory Questions"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":5126}},{"id":"sub-10-2-4","type":"subtopic","position":{"x":892,"y":5205},"data":{"label":"Coding Interviews","description":"Python · DSA"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":5205}},{"id":"sub-10-2-5","type":"subtopic","position":{"x":1152,"y":5205},"data":{"label":"Take-Home Projects"},"width":248,"height":67,"style":{"width":248,"height":67},"measured":{"width":248,"height":67},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":5205}},{"id":"topic-10-3","type":"topic","position":{"x":892,"y":5300},"data":{"label":"Staying Current","number":"10.3"},"width":768,"height":56,"style":{"width":768,"height":56},"measured":{"width":768,"height":56},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":5300}},{"id":"sub-10-3-1","type":"subtopic","position":{"x":892,"y":5368},"data":{"label":"Reading Papers","description":"arXiv · Hugging Face Papers"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":892,"y":5368}},{"id":"sub-10-3-2","type":"subtopic","position":{"x":1152,"y":5368},"data":{"label":"Tracking Model Releases","description":"Model Cards · Leaderboards"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1152,"y":5368}},{"id":"sub-10-3-3","type":"subtopic","position":{"x":1412,"y":5368},"data":{"label":"Communities","description":"Hugging Face · Meetups · Conferences"},"width":248,"height":85,"style":{"width":248,"height":85},"measured":{"width":248,"height":85},"zIndex":999,"selected":false,"selectable":true,"focusable":true,"dragging":false,"resizing":false,"positionAbsolute":{"x":1412,"y":5368}}],"edges":[]}
01 Prerequisites Build solid programming, math, and classical ML foundations for working with LLMs.
02 Deep Learning & NLP Fundamentals Learn neural network training and the classic NLP ideas that led to Transformers.
03 The Transformer Architecture Understand every Transformer component, its main variants, and the tooling to use them.
04 LLM Training & Fine-Tuning Learn how LLMs are pretrained, aligned, and efficiently adapted to new tasks.
05 Building LLM Applications Prompt models effectively and ground them in your data with robust RAG pipelines.
06 AI Agents & Evaluation Build tool-using agents and measure LLM system quality with evals and tracing.
07 Security, Safety & Governance Defend LLM systems against attacks and ship them responsibly and compliantly.
08 Deployment, Inference & Optimization Serve LLMs fast and cost-effectively on GPUs, the cloud, and local machines.
09 Multimodal AI & Emerging Trends Extend LLM skills to vision, speech, and media generation, and track where the field is heading.
10 Capstone & Career Ship portfolio-grade LLM projects and prepare for GenAI engineering interviews.
Generative AI & LLM Engineer
For developers who want to build production LLM systems. You will be able to fine-tune models, build RAG pipelines and agents, evaluate and secure them, and serve them efficiently at scale.
10 STAGES · 35 TOPICS · 151 SUBTOPICS
1.1 Programming Fundamentals
Python Mastery Generators · Decorators · Typing
Concurrency & Async Programming asyncio · Threads · Processes
Object-Oriented Programming (OOP)
Data Structures & Algorithms
HTTP APIs REST · JSON · Streaming
Version Control Git · GitHub
Command Line & Bash Scripting
1.2 Mathematical Foundations
Linear Algebra Vectors · Matrices · Tensors
Multivariable Calculus Derivatives · Gradients · Chain Rule
Probability & Statistics Distributions · Bayes' Theorem · Sampling
Information Theory Entropy · Cross-Entropy · KL Divergence
Optimization Basics Gradient Descent · Convexity
1.3 Traditional Machine Learning
Supervised vs Unsupervised Learning
scikit-learn Framework
Evaluation Metrics Accuracy · Precision · Recall · F1
Overfitting & Regularization Bias-Variance · L1 · L2 · Dropout
2.1 Deep Learning Basics
Artificial Neural Networks (ANNs)
Backpropagation & Loss Functions
Activation Functions ReLU · GELU · SiLU
Optimizers & Schedulers AdamW · Warmup · Cosine Annealing
Deep Learning Frameworks PyTorch · JAX
2.2 Classical NLP
Text Preprocessing Word Tokenization · Stemming · Lemmatization
Bag of Words & TF-IDF
Word Embeddings Word2Vec · GloVe · FastText
Subword Tokenization BPE · WordPiece · SentencePiece · tiktoken
2.3 Sequence Models
Recurrent Neural Networks (RNNs) & LSTMs
Encoder-Decoder Architecture (Seq2Seq)
Attention Mechanism Additive · Multiplicative
3.1 Transformer Anatomy
Self-Attention & Multi-Head Attention
Positional Encoding Absolute · Relative · RoPE · ALiBi
Layer Normalization Post-Norm · Pre-Norm · RMSNorm
Feed-Forward Layers MLP · SwiGLU
Efficient Attention MQA · GQA · MLA · FlashAttention
3.2 Transformer Variants
Encoder-Only Models BERT · RoBERTa · ModernBERT
Decoder-Only Models GPT · Llama · Mistral · Qwen · DeepSeek
Encoder-Decoder Models T5 · BART
Mixture of Experts (MoE)
3.3 Alternative Architectures
State Space Models Mamba · Mamba-2
Hybrid Architectures (Jamba)
RWKV & Linear Attention
3.4 Hugging Face Ecosystem
Transformers Library Pipelines · AutoModel · Tokenizers
Datasets Library
Model Hubs Hugging Face Hub · ModelScope
4.1 Pre-training
Pre-training Objectives Next-Token Prediction · Masked LM
Data Curation Filtering · Deduplication · PII Scrubbing
Scaling Laws Chinchilla · Compute-Optimal Training
Distributed Training Data · Tensor · Pipeline Parallelism · FSDP · DeepSpeed
4.2 Post-Training & Alignment
Instruction Tuning SFT · Chat Templates
Synthetic Data Self-Instruct · Evol-Instruct · Distillation
RLHF Reward Modeling · PPO
Preference Optimization DPO · ORPO · KTO · SimPO
Constitutional AI & RLAIF
Reasoning Models GRPO · Verifiable Rewards · Test-Time Compute
4.3 Parameter-Efficient Fine-Tuning (PEFT)
Low-Rank Adaptation (LoRA)
LoRA Variants QLoRA · DoRA · PiSSA
Prompt Tuning & Prefix Tuning
Adapter Layers
Fine-Tuning Tools PEFT · TRL · Unsloth · Axolotl
5.1 Prompt Engineering
Zero-Shot & Few-Shot Prompting
Chain-of-Thought (CoT)
Tree & Graph of Thoughts ToT · GoT
System Prompts & Personas
Structured Outputs JSON Mode · JSON Schema · Pydantic
LLM Provider APIs OpenAI · Anthropic · Gemini
5.2 Frameworks & Orchestration
LangChain Runnables · Prompts · Output Parsers
LlamaIndex Data Connectors · Indices · Query Engines
Haystack Pipelines · Components
DSPy Signatures · Modules · Optimizers
5.3 RAG Fundamentals
Naive vs Advanced RAG
Document Loading & Parsing PDF · HTML · Docling · Unstructured
Chunking Strategies Fixed-Size · Recursive · Semantic · Agentic
Embedding Models OpenAI · BGE · Nomic · Voyage · Cohere
Vector Databases Pinecone · Milvus · Qdrant · Chroma · pgvector
Retrieval Strategies BM25 · Dense Retrieval · Hybrid Search
5.4 Advanced RAG
Query Transformation HyDE · Multi-Query · Semantic Routing
Reranking Cross-Encoders · Cohere Rerank
GraphRAG & Knowledge Graphs
Adaptive RAG Self-RAG · CRAG · Agentic RAG
Context Compression (LLMLingua)
6.1 Agent Fundamentals
ReAct Pattern Reason · Act · Observe
Function Calling & Tool Use
Model Context Protocol (MCP)
Plan-and-Execute Architectures
Agent Memory Short-Term · Long-Term · Entity
6.2 Multi-Agent Systems
Agent Frameworks LangGraph · CrewAI · OpenAI Agents SDK · Microsoft Agent Framework
Multi-Agent Patterns Supervisor · Hierarchical · Handoffs
Execution Sandboxes Code Interpreters · Containers
Human-in-the-Loop
6.3 Evaluation
Benchmarks & Leaderboards MMLU · HumanEval · MT-Bench · LMArena
Text Metrics BLEU · ROUGE · BERTScore
LLM-as-a-Judge G-Eval · Prometheus · Pairwise Grading
RAG Evaluation RAGAS · TruLens · DeepEval
Agent Evaluation Trajectories · Tool-Call Accuracy
6.4 Observability
Tracing LangSmith · Langfuse · Phoenix
Cost & Latency Monitoring Tokens · TTFT · Throughput
A/B Testing & Shadow Deployments
7.1 Model Safety & Ethics
Bias, Fairness & Toxicity
Hallucinations & Grounding
Red Teaming Garak · PyRIT · promptfoo
Machine Unlearning & Copyright
7.2 Attacks & Defenses
Prompt Injection Direct · Indirect
Jailbreaking
Data Poisoning
Input Validation & Defensive Prompting
Guardrails NeMo Guardrails · Llama Guard · Guardrails AI
Agent Security Least Privilege · Tool Permissions
7.3 Governance & Compliance
Enterprise Compliance SOC 2 · GDPR · HIPAA
AI Regulation EU AI Act · NIST AI RMF
Documentation Model Cards · Datasheets
8.1 GPU & Compute Fundamentals
GPU Architecture CUDA Cores · Tensor Cores · HBM
Memory Hierarchy & Bandwidth
Compute-Bound vs Memory-Bound Workloads
VRAM Estimation Weights · KV Cache · Activations
8.2 Optimization Techniques
Quantization GPTQ · AWQ · GGUF · FP8 · SmoothQuant
Continuous Batching & PagedAttention
KV Cache Optimization & Prompt Caching
Speculative Decoding Draft Models · Medusa · EAGLE
Context Length Extension YaRN · RoPE Scaling
Pruning & Distillation
8.3 Serving Engines
High-Throughput Serving vLLM · SGLang
NVIDIA Stack TensorRT-LLM · Triton · NIM
Local Inference Ollama · llama.cpp · MLX · LM Studio
8.4 Cloud & Production Deployment
Managed Platforms AWS Bedrock · Azure AI Foundry · Vertex AI
Serverless GPUs RunPod · Modal · Replicate · Baseten
API Design Streaming · Rate Limiting · Caching
LLM Gateways LiteLLM · Model Routing · Fallbacks
9.1 Vision-Language Models
Contrastive Image-Text Models CLIP · SigLIP
Multimodal LLMs LLaVA · Qwen-VL · InternVL
Document AI LayoutLM · OCR-Free Parsing
9.2 Speech & Audio
Speech-to-Text (Whisper)
Text-to-Speech (TTS)
Realtime Voice Agents
9.3 Image & Video Generation
Image Generation Stable Diffusion · Flux · Midjourney
Video Generation Sora · Veo
9.4 Emerging Trends
Small Language Models & Edge AI
Continual Learning in LLMs
Autonomous Coding Agents Claude Code · Codex · Devin
Computer-Use Agents
10.1 Portfolio Projects
Production RAG Assistant
Multi-Agent Workflow
Fine-Tuned Domain Model LoRA · Evaluation · Deployment
LLM Evaluation Harness
Open Source Contributions
10.2 Job Preparation
Resume & Portfolio GitHub · Demos · Write-Ups
LLM System Design Interviews
Transformer & LLM Theory Questions
Coding Interviews Python · DSA
Take-Home Projects
10.3 Staying Current
Reading Papers arXiv · Hugging Face Papers
Tracking Model Releases Model Cards · Leaderboards
Communities Hugging Face · Meetups · Conferences