Vector Search Training
Learn vector search, hybrid search, LLMs, and RAG with a focus on accuracy, cost, reliability, scalability, and privacy.
This deep dive teaches you how to get a vector/hybrid solution with RAG to production effectively.
We'll show you how choose the right model, how to size and tune a vector/hybrid search cluster and experience, and how to scale. Afterwards, we dive into applying hybrid search to RAG with a generative model, focussing on correctness and safety.
Look no further, this course is for you.
Agenda
We train during an 8-hour hands-on workshop.
Part 1: Introduction to Vector Search
- How vector search works
- Choosing the right embedding model
- Benefits, drawbacks, and gotchas
Part 2: Hybrid Search
- Blending vector search with full text search
- Tuning for relevance
- Estimating costs
Part 3: Getting to Production
- Right-sizing your search cluster
- Tradoffs of accuracy and cost
- How hybrid search impacts SearchOps
Part 4: Retrieval Augmented Generation and LLM Enrichment
- RAG prompting and parsing
- Correctness and security practices
- To host or not to host? (OpenAI or LLama)