L10 XP00
Course 04 · Intermediate

Agentic AI Engineering

The systems track: everything around the model that makes AI products actually work.

From tokens to production: LLM fundamentals, RAG and context engineering, agent architectures with MCP and A2A, LangGraph orchestration, LoRA finetuning and local models with Ollama, evals and observability, and the security layer — injection defense, guardrails, red teaming, governance.

25 lessons6 modules~16 hours0 / 25 done · 0%
Free · no signup · progress saved in your browser

Module 01 · GenAI Building Blocks

Tokens, prompting, function calling, structured outputs, the modern stack — and the latency/cost/reliability budgets that shape everything.

  1. How LLMs Actually Workconcept10 min
  2. Prompting That Survives Contactconcept12 min
  3. Function Calling & Structured Outputswalkthrough14 min
  4. The Modern AI Stackconcept10 min
  5. Latency, Cost & Reliability Budgetsconcept11 min

Module 02 · Grounding: RAG & Context Engineering

Curate the window, anchor answers in your own data, and know which advanced retrieval patterns earn their complexity.

  1. Context Engineeringconcept12 min
  2. RAG Fundamentalsconcept13 min
  3. Build a RAG Pipelinewalkthrough18 min
  4. Advanced RAG Patternsconcept13 min

Module 03 · The Agentic Leap

From workflows to agents: architecture patterns, MCP for tools, A2A for peers, and LangGraph when orchestration gets real.

  1. What Makes a System Agenticsandbox12 min
  2. Agentic Architecture Patternsconcept13 min
  3. MCP — Model Context Protocolwalkthrough15 min
  4. A2A — Agents Talking to Agentsconcept10 min
  5. Orchestration: LangChain & LangGraphwalkthrough16 min

Module 04 · Finetuning & Local Models

When training pays off, LoRA/QLoRA in practice, and running open-weight models locally with Ollama.

  1. Finetuning: When It Pays Offconcept12 min
  2. LoRA & QLoRAconcept12 min
  3. Running Models Locally with Ollamawalkthrough14 min

Module 05 · Evals, Observability & Monitoring

Golden datasets, failure analysis, LLM-as-judge, tracing, and the production eval loop that catches regressions before users do.

  1. Why Traditional Testing Fails for AIconcept12 min
  2. LLM-as-a-Judgewalkthrough15 min
  3. Observability & Monitoringconcept12 min
  4. Production Eval Loopsconcept12 min

Module 06 · AI Security & Safety

The threat surface, injection and PII defense, guardrails that fail closed, red teaming, and governance — production readiness, proven.

  1. The AI Threat Surfaceconcept12 min
  2. Prompt Injection, Hallucinations & PIIwalkthrough15 min
  3. Guardrails & Runtime Checksconcept12 min
  4. Red Teaming, Governance & Production Readinessconcept13 min