🧠AI & Machine Learning Systems
High-throughput vector search, low-latency embedding retrieval, and resilient model inference pipelines under dynamic token generation workloads.
Foundational Invariants & Core Principles:
Stage 1: Vector Search
Sequential learning progression covering core mechanics of vector search.
Vector Embeddings, Dot Product & Cosine Similarity
Vector Embeddings, Dot Product & Cosine Similarity provides core foundational patterns in AI & Machine Learning Engineering. Represent unstructured text and audio as high-dimensional mathematical vectors and measure semantic similarity.
Embedding Model Selection
Embedding Model Selection is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Vector Databases, HNSW Indexing & RAG Pipelines
Vector databases index high-dimensional vector embeddings (e.g. 1536-dim vectors) produced by AI models, enabling sub-millisecond semantic search via Hierarchical Navigable Small World (HNSW) graph indexing.
Hybrid Vector & Keyword Search
Hybrid Vector & Keyword Search is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
RAG Retrieval Pipelines
RAG Retrieval Pipelines is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Stage 2: RAG Architecture
Sequential learning progression covering core mechanics of rag architecture.
Context Compression for RAG
Context Compression for RAG is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Deep Learning Foundations: Backpropagation, Transformers & Optimizers
Deep Learning Foundations: Backpropagation, Transformers & Optimizers provides core foundational patterns in AI & Machine Learning Engineering. Derive backpropagation matrix calculus, Softmax activation, AdamW optimization, and dense Transformer neural networks.
Transformer Self-Attention
Transformer Self-Attention is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
LLM Quantization, GPTQ/AWQ & Inference Performance
LLM Quantization, GPTQ/AWQ & Inference Performance provides core foundational patterns in AI & Machine Learning Engineering. Compress large language model weights from FP16 to INT4/INT8 to optimize GPU VRAM memory bandwidth and token throughput.
Model Serving & GPU Batching
Model Serving & GPU Batching is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Stage 3: Embedding Pipelines
Sequential learning progression covering core mechanics of embedding pipelines.
Semantic Caching for LLMs
Semantic Caching for LLMs is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
LLM Fine-Tuning & LoRA
LLM Fine-Tuning & LoRA is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Prompt Engineering, LLM Reasoning & Thinking Models
Prompt Engineering, LLM Reasoning & Thinking Models provides core foundational patterns in AI & Machine Learning Engineering. Design structured prompts, Few-Shot examples, Chain-of-Thought reasoning, and mitigation for LLM hallucinations.
Prompt Injection Defenses
Prompt Injection Defenses is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
LLM Agent Tool Use
LLM Agent Tool Use is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Stage 4: LLM Caching
Sequential learning progression covering core mechanics of llm caching.
AI Agents, LangGraph & Stateful Tool Orchestration
AI Agents, LangGraph & Stateful Tool Orchestration provides core foundational patterns in AI & Machine Learning Engineering. Orchestrate autonomous agentic workflows using state machines, function calling contracts, and multi-agent loops.
Multi-Agent Orchestration
Multi-Agent Orchestration is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
LLM Evaluation Harnesses
LLM Evaluation Harnesses is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Feature Stores for ML
Feature Stores for ML is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
ML Pipeline Orchestration
ML Pipeline Orchestration is a foundational topic in AI & Machine Learning. Engineers use it to make predictable trade-offs under load, failure, and evolving requirements.
Stage 5: Inference Serving
Sequential learning progression covering core mechanics of inference serving.
Explore Other Domain Curriculums
Distributed Systems
High-availability replication, event loops, quorum consensus, and horizontal database sharding.
Software Architecture
Decoupled event streaming, microservice resilience, CQRS, and domain-driven design.
Design Patterns
GoF structural/behavioral patterns, circuit breakers, sagas, and transactional outbox patterns.
Software Engineering
Clean code craft, automated testing pyramids, CI/CD pipelines, refactoring, and technical debt.