Skip to content

🗺️ Course Roadmap

Course Overview

  • Duration: 12–16 weeks core + 4-6 weeks electives
  • Target: Mid-level software engineers
  • Style: Text-first with embedded code

📚 Core Curriculum

Phase 0: Foundations & Mindset (1–2 weeks)

  • 0-1 Why AI Agents? - Evolution, automation vs autonomy
  • 0-2 Core Concepts - Planner ∙ Executor ∙ Retriever loop
  • 0-3 Thinking Like an Agent - Problem decomposition, behavior trees

Phase 1: Essential Tools & Techniques (2–3 weeks)

  • 1-1 Python for Agents - Tool functions, async HTTP, Pydantic
  • 1-2 Prompt Engineering - Zero/few-shot & CoT prompts
  • 1-3 Tool Wrappers & LangChain - Chains vs agents, custom tools
  • 1-4 Memory & Basic RAG - Vector stores, embeddings
  • 1-5 Model APIs - GPT-4o, Claude 4, OSS models

Phase 2: Agentic Workflows (2–3 weeks)

  • 2-1 Agent Loop Deep Dive - ReAct internals, dead-loop guards
  • 2-2 Prompt & Tool Chaining - Conditional branches, fallbacks
  • 2-3 Hybrid RAG & Context - Vector vs graph vs hybrid
  • 2-4 Evaluation & Tracing - Cost/latency metrics, LangSmith

Phase 3: Multi-Agent Orchestration (2–3 weeks)

  • 3-1 Multi-Agent Orchestration - Framework survey
  • 3-2 Advanced RAG Pipelines - Multi-tenant stores, scaling
  • 3-3 State & DAG Orchestration - Event-driven vs static DAGs
  • 3-4 Collaboration Patterns - Planner-worker-reviewer
  • 3-5 Framework Survey - CrewAI / LangGraph / AutoGen

Phase 4: Production Deployment (2–3 weeks)

  • 4-1 Production Deployment - Containerization, serverless
  • 4-2 Containerization & Serverless - Docker, AWS Lambda
  • 4-3 CI/CD & Secrets - GitHub Actions, secret scanning
  • 4-4 Observability & Incidents - Logs, SLO/SLA, runbooks
  • 4-5 Security & Compliance - OAuth, RBAC, sandboxing

Phase 5: Advanced Safety (2–3 weeks)

  • 5-1 Advanced Safety - Multi-layer guardrails
  • 5-2 Guardrails Engineering - Safety frameworks
  • 5-3 RLHF & Policy Tuning - DPO, reward models
  • 5-4 Continuous Evaluation - REALM-Bench, CI benchmarks

Phase 6: Portfolio & Leadership (2–3 weeks)

  • 6-1 Portfolio Leadership - Portfolio development
  • 6-2 Capstone Showcase - End-to-end agent project
  • 6-3 Open Source & Community - Repo structure
  • 6-4 Lifelong Learning Loop - Research feeds, trends

🎓 Elective Tracks (4-6 weeks each)

  • 🌍 AI for Geospatial & Green Infrastructure
  • 💰 Financial Agents & Compliance
  • 🏥 Healthcare Agents & Voice Biomarkers
  • ⚡ Low-Level LLM (C++/CUDA)
  • 🎤 Voice Agent Engineering
  • 🏢 Enterprise Orchestration (Azure/Power Platform)
  • 👁️ Multimodal Agents (Vision + Audio)

📊 Course Stats

  • Core Phases: 6
  • Core Lessons: 22
  • Elective Tracks: 7
  • Total Duration: 12–16 weeks + electives
  • Content: ≈ 39,000 words, ≈ 40 images/diagrams

Ready to Start?

🚀 Begin Learning