This course is popular
3 people enrolled last week.
Improve RAG with smarter agents and better retrieval
Retrieval is the AI industy's big blindspot. Poor retrieval steers agents astray - down rabbit holes, with false assumptions. We take AI-driven retrieval head on in this course.
🧑🏻🏫 I've taught hundreds of practitioners at places like Meta, Apple, Capital One, Wikipedia, and more how to ship better retrieval / RAG
🤝 You'll get exclusive access to our cheat at search community. With hundreds of professionals solving tough problems, ongoing updates, and discounted access to future cohorts.
We'll cover:
How do we know when agents have been led astray? How can we evaluate retrieval? What does good search look like for an agent?
Agentic Search vs Classic RAG: how do agents interact with search in 2026? What does the modern retrieval stack look like?
Beyond embeddings - What ranking / retrieval techniques do state of the art search teams deploy to improve search
Autoresearching to smarter search: Use + build a coding agent that optimizes retrieval code like an ML model
The major use-cases: From agentic memory, to your company's knowledge base
You'll leave knowing how to tackle your AI teams toughest search problems
What you’ll learn
Rethink search to put tireless agents at work for human users
Build search backends agents can reason about
Why agents love simplicity over complexity
Focusing tools on domain tasks, not "search"
Sidestep complex NLP with jev and LLMs to better understand queries
Iterate on query understanding through agentic approaches
Organize and structure content to be retrievable by an agent
How to integrate agentic approaches into a traditional search stack
Best practices to manage the agentic loop to save tokens, money, and prevent context rot
Technical approaches to extracting agentic insights (ie code generation)
An agent will by default use its judgments to iterate on relevance on its own. But shouldn't it use humans?
Integrating external feedback (user clicks, etc) into agentic relevance
Beyond LLM as a Judge to exploring how to iterate and improve retrieval for agents and humans
Who this course is for
AI Engineers - maybe you've hacked a bit on some simple embedding based RAG and feel stuck. Or you're just getting started with search.
Search technologists that want to learn how to apply agents and LLMs to their work
Team leads and PMs that want to consider how agentic / LLM / RAG based search can be improved
Prerequisites
Python coding
We use jupyter notebooks in Python for simple examples. It's nice to have an appreciation for the standard python stack (numpy, etc)
Agent / AI basics
It's a good idea to have interacted with, for example, OpenAI's API at a basic level
What's included
Live sessions
Learn directly from Doug Turnbull in a real-time, interactive format.
Lifetime access
Go back to course content and recordings whenever you need to.
Community of peers
Stay accountable and share insights with like-minded professionals.
Certificate of completion
Share your new skills with your employer or on LinkedIn.
Maven Guarantee
Your purchase is backed by the Maven Guarantee.
Course syllabus
Week 1
Oct
5
Lesson 1 - Information Retrieval as an agentic process
Oct
6
Max Rumpf - Agentic Search Models w/ Sid.AI (Guest Talk)
Oct
7
Lesson 2 - LLM Query Understanding
Oct
9
Office Hours
Week 2
Oct
12
Lesson 3 - An agentic harness - that searches
Oct
13
Daniel Svonava - serving inference for search+agent use cases (Guest Talk)
Oct
14
Lesson 4 - Beyond LLM as a judge
Oct
16
Office Hours
Free resource
Cheat at Search Essentials: Vectors and Embeddings
Learn how embeddings implement 'semantic search'
What is a 'vector embedding' and why do they capture meaning? And how can that be used to build a search system?
Tackle the practical realities vector retrieval
Why do we need a vector database? And why are they different from traditional search engines? Should they be different?
How are vector + lexical techniques combined?
Offer a concrete and concise explanation of how you will help students understand and apply this lesson.
Schedule
Live sessions
2-4 hrs / week
Three hours of courses. Optional guest talks from industry experts every week. Office hours to deep dive into your problems with your instructor
Mon, Oct 5
5:00 PM—6:30 PM (UTC)
Tue, Oct 6
5:00 PM—6:00 PM (UTC)
Wed, Oct 7
5:00 PM—6:30 PM (UTC)
Projects
1-2 hrs / week
Optional labs to expand your knowledge and deliver into your team's codebase
Why agents completely change search+RAG
Walking through exactly why / how agents change the game for search+RAG
How to eval retrieval for RAG
Hamel Hussain + Doug Turnbull discuss RAG evaluation
Frequently asked questions
Maven for Teams
Reimbursement
Get your company to pay
Everything L&D needs: email template, receipts, and certificate of completion.
Get reimbursedPrivate cohort
Run a cohort for your org
A dedicated cohort with a custom schedule and curriculum, tailored to your team.
Book a private cohort