Architecting Vector Search: Performance, Scaling, and Continuous Evaluation

Vector search has shifted from the "new shiny technology" for RAG applications in 2024 to a foundational part of every second AI company's stack.

With the loss of hype came great responsibility (and actual usability). That usability implies many architectural considerations. How do you build search systems that are reliable and performant? How do you ensure that the pipeline keeps yielding high-quality search results? And, overall, how do you continuously evaluate the system to measure the impact of every change?

This masterclass will attempt to answer these questions with a hands-on approach that guides you from a PoC running on your laptop to a production app deployed to a Kubernetes cluster.

Who Should Attend

Anyone working with vector search and wanting to bring it to production.

What You'll Learn

• Vector search concepts (data processing, embedding, indexing, search evaluation metrics)
• How to measure search quality and performance
• How to deploy and scale vector search from PoC to production
• How to keep vector search systems reliable and performant in the long term

What to Bring

Docker with Docker Compose, and kind. Windows users will also need WSL2.

Your trainers

Clelia Astra Bertelli

Clelia is a DevRel Engineer at Qdrant and previously worked as Open Source Engineer at LlamaIndex. Passionate about all things OSS, she's currently focusing on AI search performance, scalability and infrastructure. In her free time, she likes coding in Python, Go or Rust, hiking and watching TV series.

LinkedIn · X · GitHub

Evgeniya "Jenny" Sukhodolskaya

Developer Advocate at Qdrant with 9 years of IT experience across software engineering, machine learning, and developer advocacy. Holds a master's degree in Data Analytics and Engineering from the Technical University of Munich. Passionate about NLP and Information Retrieval. Believes in conference-, complaints-, and memes-driven development :)

LinkedIn · X · GitHub