An AI engineering course in eight modules, from generative AI basics and the first LLM call to agents that run on their own. Instructor-led, on site or live online, with no prior AI experience needed.
When one person holds the AI knowledge, every feature queues behind them. This AI engineering course trains everyone who reads, writes, tests or runs the code. No prior AI experience is assumed and nobody is asked to leave their stack behind — module one starts with the TypeScript patterns the rest of the program is built on.
Backend, front-end and full-stack developers putting generative AI and LLMs inside real applications, and platform engineers who have to keep them running.
The people who test, evaluate, secure and deploy what gets built — hallucination and bias detection, evaluation metrics and deployment are modules of their own.
Leads and architects accountable for what ships, who need the same vocabulary as the team reviewing and approving the work.
Most engineers can call a generative AI API. Far fewer can ship retrieval that holds up on real data. Each module of this AI engineering course builds on the last and closes with a hands‑on project.
TypeScript essentials for AI work — types, interfaces, generics, async and streaming patterns, Zod validation, and AI backends on Node and Express.
Learning objectiveStay in TypeScript throughoutOne language from the first module to the capstone, so nothing is rewritten for production.
Certify Your TeamHow generative AI and LLMs actually work — tokenisation, context windows, model selection, prompt engineering, embeddings, and a first RAG pipeline.
Learning objectiveKnow when not to use a modelSpot hallucination failure modes and token cost traps before users do.
Certify Your TeamProduction integration across OpenAI, Anthropic, Gemini and the Vercel AI SDK — streaming, function calling, tool use, error handling and moderation.
Learning objectiveIntegrate LLM APIs in productionPut a model into a live feature and switch providers without rewriting it.
Certify Your TeamRAG end to end — document chunking, vector databases including pgvector, Pinecone, Weaviate and Qdrant, LangChain stores and Mastra.
Learning objectiveBuild a RAG pipeline end to endGround answers in your own documents, inside a working application.
Certify Your TeamAgents with Mastra, VoltAgent and LLM-EXE — tool use, memory, multi-provider APIs and Zod‑typed workflows.
Learning objectiveShip agents with tools and memoryAgents that finish real tasks on their own, not demos.
Certify Your TeamFull applications with React, Node, Express and LangChain — chat history, vector retrieval, NestJS backends and deployment.
Learning objectiveWork to production standardsRetry logic, cost estimation and graceful failure — the parts that only matter when something breaks.
Certify Your TeamBias and hallucination detection, evaluation metrics, EU AI Act compliance, multimodal search and security pipelines for production.
Learning objectiveApply evaluation, security and the EU AI ActShow that a system is accurate, fair, secure and compliant before it goes live.
Certify Your TeamBuild and present a complete application — a RAG question-answering system, an autonomous agent, a multimodal app or a domain chatbot.
Learning objectiveKeep a deployed capstonePresented live, left running after the AI engineering course ends and owned by the team.
Certify Your TeamEvery concept, project and line of code is TypeScript. The language of production AI engineering, from day one.
Each module closes with a working project. Token counters, RAG pipelines, deployed chatbots — running code the team owns.
Error handling, retry logic, cost estimation, moderation and security are treated as first-class, not as an appendix.
Work across OpenAI, Anthropic, Gemini, Mastra and LangChain rather than locking the team to a single vendor.
Generative AI failure cases and hallucination risk are taught next to the capabilities. Engineering judgement needs both.
Responsible AI is not a bolt-on module. The EU AI Act, bias detection and security run from the foundations to the capstone.
Every module closes with a hands-on project that is reviewed while the program runs. Those eight projects are the codebases the team keeps.
A complete application built and presented live — a RAG question-answering system, an agent, a multimodal app or a domain chatbot. It is deployed and it stays.
A Gen AI Foundations Certificate per participant who completes the AI engineering course and passes the capstone, and a direct route into Architecture.
Training that ends with working software rather than better prompts. This program runs eight modules from the first LLM call through generative AI, retrieval and tool use to agents that complete tasks on their own, all in TypeScript.
15+ hours across eight modules, instructor-led, on site at your offices or live online. Dates are agreed with you and the group runs together.
Engineers who already ship software professionally and now have to put LLMs and agents into a real product. No prior AI work is assumed.
No. The whole AI engineering course is in TypeScript, and no Python is needed at any point.
A deployed capstone each engineer keeps, the patterns behind it, and the Gen AI Foundations Certificate for those who pass the assessments.
The AI engineering course is built for the group and priced on team size, level and timeline.
Design agentic systems that survive production — hybrid retrieval, knowledge graphs, multimodal pipelines and agents that coordinate.
Learn More →Both engineering tracks as one path — from TypeScript fundamentals through to autonomous systems, with two capstones.
Learn More →21 hours building production multi-agent systems — anatomy, orchestration, guardrails and a launch review.
Learn More →Organisations start with a conversation about where they stand. Professionals start with a bootcamp or a summit.