Build your own AI agent in 30 hours. Go from "what is an agent?" to a working, tool-using, memory-equipped agent you can demo, deploy, and put in your portfolio — no PhD required.
Every concept is anchored to one outcome: a real agent that runs.
An autonomous assistant that reasons, picks tools, remembers context, and completes multi-step tasks on its own.
The perceive → reason → act → observe loop that powers every agent — built once from scratch, then with frameworks.
Tools, memory, RAG, guardrails, evaluation, cost control and deployment — what separates demos from real products.
A recognised Jetking Certification on completion — plus a documented, demoable capstone agent for your portfolio.
Hands-on with what teams use today: CrewAI, LangGraph, OpenAI Agents SDK, and MCP.
When an agent is the right tool — and when a simple script or single prompt beats it. Real engineering judgement.
Designed for Jetking students & alumni stepping into the agentic era.
5 weekends · Saturdays & Sundays · 3 hours per session · 10 sessions = 30 hours.
Tap any module to expand. Black tags mark a build milestone on your capstone agent.
Get the lay of the land and a working environment, so you never get stuck on setup later. Meet the agent you'll build by the end.
Draw the line between a chatbot, an automation, and a true agent. Understand autonomy, goals and the agent loop — the backbone of everything that follows.
Learn exactly as much about large language models as an agent-builder needs — no more, no less — and write prompts that behave reliably.
Build a minimal agent from raw API calls — no framework — so you understand the plumbing every framework hides. Implement the Reason+Act (ReAct) pattern by hand.
Give your agent hands. Define tools, wire up function calling, and let the model decide which tool to use and when.
Give your agent a memory and a knowledge base. Use embeddings, vector stores and retrieval-augmented generation so it can answer from your own documents.
Move from single steps to goals. Teach your agent to break a task into sub-tasks, check its own work, and recover from mistakes.
Rebuild your agent the fast way. Tour today's leading frameworks, then re-implement your agent in CrewAI and meet LangGraph & the OpenAI Agents SDK.
One agent is good; a team is powerful. Coordinate specialised agents with roles, handoffs and supervisor patterns.
Plug your agent into real apps and data using the Model Context Protocol (MCP) — the standard way agents reach email, files, databases and your own services.
Make your agent trustworthy and affordable. Handle hallucination and prompt injection, add human-in-the-loop checks, measure quality, and control your token bill.
Wrap your agent in a simple interface, deploy it, and present it. Finish with a documented capstone agent ready for your portfolio.
Current as of 2026. Free / low-cost tiers throughout.
You don't wait until the end to build. From Weekend 2 onward, every milestone adds a real capability to your agent. By the final Sunday you've assembled a complete, autonomous, tool-using assistant — and you understand every piece of it.
Bare-metal agent loop running
Agent uses tools & APIs
Memory + your own knowledge
Rebuilt on a framework
Connected, safe, deployed
One payment. Lifetime access. No hidden fees.
Tap a question to expand.
No prior AI or data-science background needed. You should be comfortable with a computer and willing to write a little Python — we teach the just-enough code, step by step, in every session.
Every session is recorded and yours for life. You can catch up anytime and still complete the capstone. The private community is there for any doubts along the way.
Yes — you earn a Jetking Certification on completing the course and submitting your capstone agent. The project itself is portfolio-ready to show employers and clients.
A laptop and a stable internet connection. Every tool used has a free or low-cost tier, and we show you how to stay within it. You'll set up your own free accounts and API keys in the first session.
The course fee covers the teaching, recordings, materials and certificate. LLM API usage during the course is minimal and runs on free or low-cost tiers using your own keys — we keep your spend tiny and teach cost control directly.
You're covered by our 100% Money-Back Guarantee. If the course isn't for you, tell us and we'll refund your full fee — no hard feelings. See the Refund Policy for how to claim.