outpaczAgentic AI Engineering
Design, build and deploy tool-using AI agents in 7 weeks: LangGraph, MCP, memory, multi-agent systems, evals and deployment. Hands-on, live labs.
32 lessons · 31 hours · ₹15,000
What you will build
Design agents the right way
Choose between workflows and agents, and apply proven patterns: chaining, routing, orchestrator-workers
Build agents with LangGraph
Use state, routing, checkpoints and human-in-the-loop approvals
Give agents memory and context
Manage short and long-term memory, context pruning and agentic RAG
Connect tools through MCP
Build and use MCP servers, browser agents and sandboxes, with tool security in mind
Orchestrate multi-agent systems
Coordinate supervisor and sub-agents, and know when multi-agent costs more than it gains
Test, secure and deploy
Evaluate agent trajectories, red-team for prompt injection, and ship with queues and WhatsApp or Telegram integration
Curriculum
Bridge videos (prerequisites)
- Python for LLM work · 180 min
- Prompt writing basics · 90 min
- SQL and DBMS basics · 180 min
- APIs and auth · 120 min
Week 1: Agent fundamentals
- Agents vs workflows, and when not to use agents · 15 min
- Core patterns: chaining, routing, parallel, orchestrator-workers, evaluator-optimiser · 25 min
- The ReAct loop and tool-calling mechanics · 20 min
- Live lab: build an agent from scratch in plain Python · 120 min
Week 2: LangGraph and agent SDKs
- State, nodes and conditional edges · 20 min
- Checkpoints and human-in-the-loop · 20 min
- Agent SDKs and how to choose a framework · 20 min
- Live lab: research agent with citations and an approval step · 120 min
Week 3: Memory and context engineering
- Short-term vs long-term memory · 20 min
- Context engineering: summarising, pruning, notes files · 20 min
- Agentic RAG · 20 min
- Live lab: personal ops assistant with cross-session memory · 120 min
Week 4: MCP and tools
- MCP clients and servers explained · 15 min
- Building a custom Python MCP server · 25 min
- Ready-made servers, browser agents, sandboxes and tool security · 20 min
- Live lab: wrap an API as an MCP server · 120 min
Week 5: Multi-agent systems
- Supervisor, sub-agents and handoffs · 20 min
- Parallel research and critic loops · 15 min
- When multi-agent hurts, and business workflows in n8n · 25 min
- Live lab: multi-agent team for a business task · 120 min
Week 6: Reliability
- Agent evals: outcome and trajectory · 20 min
- Tracing, retries, budgets and guardrails · 20 min
- Prompt injection and human-approval policies · 20 min
- Live lab: eval and red-team your own agent · 120 min
Week 7: Deployment
- FastAPI and queues for long-running jobs · 20 min
- Webhooks and WhatsApp or Telegram bot integration · 20 min
- Cost control, scheduling and AWS · 20 min
- Live: capstone demo day and viva · 120 min
Syllabus
Who this course is for
Students and working professionals who want to build AI agents that use tools, remember context and work on real tasks, and who want to know when a simple workflow is the better choice.
Prerequisites
Python, APIs, DBMS and prompt writing. The bridge videos at the start cover all four, about 9.5 hours in total.
How the course runs
- Recorded lessons are released every Monday, in short videos of about an hour a week
- One live session each weekend: a 2-hour hands-on lab with mentors, plus doubt solving. Replays are available
- A problem statement every week, reviewed by mentors, with a leaderboard
- Batch community group with daily doubt-desk hours
- Live session timings are announced before each batch starts
Time commitment
About 21 hours of core sessions over 7 weeks, plus 9.5 hours of bridge videos.
Assessment and certificate
- Weekly problem statements
- Capstone: a deployed agent app with an evaluation suite
- A short mentor viva
- Verified certificate on scoring 70% or above with the capstone submitted
