Build AI applications you actually understand.
From LLMs and RAG to agents, MCP and production AI systems — learn how the pieces fit together and when to use them.
- Format
- Live · Saturdays only
- Next batch
- January 2027
- Starting from
- €10
What does our policy say about refunds?
Retrieved
The refund terms are set out in the policy document [1] and confirmed in the FAQ [2].
People who want to build AI systems, not just use them.
This is not a "prompt engineering" course. It is for engineers and developers who want to understand how to build reliable AI applications — and make sound technical decisions when doing so.
Software engineers
You can code but you want to understand how to integrate LLMs, RAG and agents into real applications properly.
Data professionals
You work with data and want to add AI Engineering capability — building systems rather than just calling APIs.
Technical leads & architects
You need to evaluate AI architecture decisions, understand trade-offs, and guide your team with confidence.
Students & career changers
You want to build practical AI Engineering capability — with real projects — not just a certificate.
Understanding before implementation.
We don't start with "call the API and see what happens." We start with why — and build from there.
Model behaviour, not just model usage
You will understand tokens, context windows, hallucination causes and model selection trade-offs — so you can engineer around limitations rather than be surprised by them.
Architecture decisions, not pattern matching
When do you use RAG? When do you add an agent? When does MCP make sense? These are engineering decisions — and we teach you to reason about them, not just copy patterns.
Reliability is part of the curriculum
Evaluation, tracing, testing, fallbacks and observability are not optional extras. They are how AI Engineers ship systems that work in production — not just in demos.
Choose how deep you want to go.
Three different entry points. One consistent teaching philosophy.
Masterclass
Understand the field before committing.
- 3-hour live session
- Practical introduction to AI Engineering
- Real-world AI application architecture
- Live demonstrations with LLMs
- Q&A with the instructor
- Ideal for exploring the field
Free for colleges & student communities
Organise a session with DataKern and the Masterclass is free for your college or student community.
Crash Course
Learn the core concepts and build something real.
- 6 live sessions (Saturdays only · 2.5 hours each)
- Practical assignments each session
- LLMs, RAG, vector databases
- AI agents, tool calling and MCP fundamentals
- End-to-end AI application project
Full Course
Build, troubleshoot, evaluate and defend real engineering systems.
- 12 live sessions (Saturdays only · 2.5 hours each)
- Two major capstone projects
- Advanced RAG evaluation and reliability
- Production AI application engineering with FastAPI
- LangGraph agent engineering in depth
- MCP architecture and enterprise integration
- Observability, security, cost and governance
- Guest lectures from industry experts
- Personal coaching session
- Portfolio & resume review
- Mock technical interview preparation
Not sure which level is right for you?
Understand the field. Explore what Data/AI Engineering actually is before committing to a full course.
Understand + Build. Core concepts and a complete real-world project. Valuable as a standalone course.
Engineer + Evaluate + Govern + Architect + Defend. Deeper engineering, production patterns, architecture decisions and personal coaching.
You can enter at any level — Masterclass is an optional starting point. Many students begin directly at Crash Course or Full Course.
What we cover across all 12 sessions.
All classes are held live on Saturdays (2.5 hours each). Sessions 1–6 form the foundation (Crash Course + Full Course), while Sessions 7–12 go deeper into production agents, evaluation, and architecture (Full Course only).
Loading curriculum...
Build applications you can explain and defend.
The Full Course includes two major project milestones. The Crash Course delivers the first.
End-to-End AI Application
Session 6 · Crash Course + Full Course
Problem
Choose a realistic scenario and build an appropriate AI application for it.
Architecture
Combine LLMs, RAG, tools, agents and MCP only where appropriate.
Technology
Python · LLM API · ChromaDB · LangChain · LangGraph · FastAPI · MLflow
Outcome
A working AI application with evaluation and tracing — built by you.
Production AI Architecture & Decision Review
Session 10 · Full Course only
Problem
Take the main project beyond a working demo and reason about production requirements.
Challenge
Compare direct LLM, RAG, agent and MCP-enabled approaches. Justify the right one.
Focus
Reliability, security, observability, cost and governance.
Outcome
An architecture decision package you present and defend.
What you'll work with.
A practical, low-friction stack chosen for what it teaches, not for trend value.
Why this AI stack?
We chose Python, LangChain/LangGraph, ChromaDB, FastAPI, and MLflow because they represent the most practical, widely-adopted tools for building production AI applications today. Every tool exists because it solves a real engineering problem — not because it's trendy.
Model-agnostic by design
We primarily use OpenAI's APIs for foundation models, but we also cover open-source alternatives and how to build system architectures that are model-agnostic. You'll learn to reason about model selection — not depend on a single provider.
What you'll be able to do.
After the Crash Course
Understand the major building blocks of modern AI applications and build an end-to-end application using an appropriate combination of LLMs, RAG, tools, agents and API delivery.
After the Full Course
Build, evaluate, troubleshoot and productionise AI applications. Reason about RAG quality, agent control, MCP integration, reliability and cost. Explain and defend engineering decisions in technical interviews.
Which format is right for you?
| Masterclass€10 | Crash Course€55 | Full Course€99 | |
|---|---|---|---|
| Duration | 3 hours | 15 hours | 30 hours |
| Live sessions | 1 session | 6 sessions (Saturdays only) | 12 sessions (Saturdays only) |
| Schedule | Single Saturday | Saturdays · 2.5h each | Saturdays · 2.5h each |
| Practical assignments | — | ✓ Each session | ✓ Each session |
| Core concepts | Introduction | ✓ Full foundation | ✓ Full foundation |
| Hands-on project | — | Main Project 1 | Projects 1 & 2 |
| Advanced engineering | — | — | ✓ Sessions 7–9 |
| Architecture | Overview | — | ✓ In depth |
| Personal coaching | — | — | ✓ Session 11 |
| Portfolio review | — | — | ✓ Session 11 |
| Interview preparation | — | — | ✓ Session 12 |
Free guides
Plan your path into AI engineering.
AI Engineer Roadmap: From Python to Production Agents
A practical AI engineer roadmap: the skills to learn, in order, from LLM basics and RAG to evaluation, agents and shipping a real application.
Read the guide · 8 min read →AI EngineeringRAG vs. Fine-Tuning
When to use which, and why the industry frequently gets it wrong when giving LLMs custom knowledge.
Read the guide · 5 min read →AI EngineeringEvaluation of LLMs
How to systematically measure the performance of AI systems and solve the
Read the guide · ~10 min read →Frequently Asked Questions
Is this a live online AI engineering course?
Is there an AI engineering course near me?
Is this a prompt engineering course?
Do I need a strong Python background?
Will we cover fine-tuning?
What APIs will we use?
Ready to start building?
Data Engineering and AI Engineering batches are in progress. Applications are open for the January 2027 batch.
Next masterclass expected mid-December / mid-January · Start risk-free before committing.