AI Application Architecture
Training vs Inference
Probabilistic Model Behavior
Model Selection and Trade-offs
Tokens and Tokenization
Context Windows
Transformers and Attention
Next-Token Prediction
Pretrained and Fine-Tuned Models
Sampling and Decoding Parameters
Prompt Anatomy and Input Formatting
System Prompts
Zero-Shot and Few-Shot Prompting
Reasoning Prompt Patterns
Structured Output Prompting
Prompt Templates and Versioning
Context Sources and Assembly
Long-Context Processing
Context Compaction
Conversation State and Memory
Context Isolation and Failure Modes
Model API Integration
Streaming Responses
Tool and Function Calling
Structured Outputs and Schema Validation
Retries, Timeouts, and Rate Limits
Model Routing and Fallbacks
Prompt Caching and Token Cost Control
Embedding Fundamentals and Model Selection
Similarity Metrics
Vector Database Fundamentals and Indexing
Metadata Filtering
Semantic and Hybrid Search
RAG Architecture
RAG vs Fine-Tuning
Document Ingestion
Chunking Strategies
Retrieval Strategies
Query Transformation and Reranking
Context Assembly, Citations, and Grounding
Access-Controlled Retrieval
RAG Evaluation and Failure Modes
Agent Architecture and Execution Loops
ReAct Agents
Tool-Using Agents
Planning and Task Decomposition
Agent State and Memory
Multi-Agent Systems
Human-in-the-Loop and Agent Guardrails
Agent Evaluation and Failure Recovery
MCP Architecture
MCP Hosts, Clients, and Servers
MCP Tools, Resources, and Prompts
MCP Data and Transport Layers
Building and Securing MCP Integrations
Fine-Tuning Decision Criteria
Training Dataset Preparation
Supervised Fine-Tuning
Parameter-Efficient Fine-Tuning and LoRA
Fine-Tuned Model Evaluation
Evaluation Dataset Design
Deterministic and Model-Based Evals
Human Evaluation and LLM-as-a-Judge
Evaluation Metrics and Regression Testing
LLM Tracing and Logging
Latency, Token, and Cost Monitoring
Production Quality Monitoring
AI Reliability, Scaling, and Release Strategies
Prompt Injection and Jailbreaks
Indirect Prompt Injection and Untrusted Content
AI Data Privacy and PII Protection
Content Moderation and Output Guardrails
Hallucination and Grounding
Bias and Fairness
Adversarial Testing and Least-Privilege Tool Access
Multimodal Model Fundamentals
Image Understanding
Image Generation
Video Understanding
Speech-to-Text
Text-to-Speech and Audio Processing
