Learn directly from Trey Grainger & Doug Turnbull

Two things stood out: the guest lectures from practitioners working on these problems at scale, and the weekly office hours where I could bring actual retrieval problems from work and get thoughtful answers.

As an AI engineer, I didn't have a foundation in information retrieval. I now understand the full landscape of IR: boosting results, incorporating recommendation, learning to rank and how to integrate these areas with embeddings. The guest lectures were fantastic and I leave with a much more complete understanding of building AI around knowledge bases. Highly recommended, especially for those with no IR knowledge

This course was really great, it has helped me define my search roadmap for the next year. The insights around clickstream have already yielded results for us. This course is a bargain for what you get.

I really enjoyed this course and found it especially valuable as someone already working in search. The sessions on multivector search and Learning to Rank (L2R) helped deepen my understanding of modern retrieval techniques and the tradeoffs involved in building production search systems. I also appreciated the balance between theory, practical implementation, and real-world examples. The course gave me several new perspectives that I can apply directly to my work. I would highly recommend it to search engineers, ML engineers, and anyone working on retrieval, RAG, or AI-powered search who wants to build a stronger technical foundation.

This course is *hugely* comprehensive in the AI-powered search domain, with high ROI. First, the lectures provided a decent overview of the 400+ page book (as much as is possible - I’ve had to reference the recordings more than once to review). Guest lectures from leading experts on semantic IDs, late interaction, etc provided exposure to SOTA methods and their applications in industry. For every session, Doug and Trey also answered many questions relevant to real life use case across industries. Second, the notebooks on the GitHub provide enough toy datasets and codes snippets to set you up to let you cook. Most of all, I really appreciate how well Trey and Doug, both being experts in their fields, are able to explain things so well, provide nicely documented notebooks, and make information accessible to even beginners in search.