RAG & AI Agents: Production Deep Dive
Build retrieval and agent systems that actually work in production.
Created by Dr. Ivan Petrov · ML Systems Engineer
30-day money-back guarantee
This course includes
- 16 video lessons
- 3h 58m of content
- Downloadable resources
- Full lifetime access
- Certificate of completion
- 30-day money-back guarantee
What you'll learn
- Design retrieval that returns the right context
- Chunking, embeddings, hybrid search and reranking
- Build agent loops with reliable tool use
- Defend against prompt injection and jailbreaks
- Build evals that catch quality regressions
- Control latency and cost in production
About this course
The gap between a RAG demo and a RAG system you can trust in production is enormous. This advanced course closes it. You'll build retrieval-augmented generation and tool-using agents the way they're built at serious companies: with real evaluation, cost control, guardrails and observability. Expect depth. We cover chunking and retrieval strategy, hybrid search and reranking, agent loops and tool design, prompt injection defence, evals that catch regressions, and how to keep latency and cost sane at scale.
Course content
4 modules · 16 lessons · 3h 58m
- Why naive RAG fails in productionPreview13m
- Embeddings and vector search, properlyPreview17m
- Chunking strategies that matter15m
- Hybrid search and reranking16m
Requirements
- Solid programming experience
- Familiarity with LLM APIs (see Build with LLMs)
Your instructor
Dr. Ivan Petrov
ML Systems Engineer
Ivan builds LLM infrastructure for a fintech serving millions of users. He's shipped RAG and agent systems that survive contact with real traffic — and has the scars to prove it.
30-day money-back guarantee
This course includes
- 16 video lessons
- 3h 58m of content
- Downloadable resources
- Full lifetime access
- Certificate of completion
- 30-day money-back guarantee