Jo Bergum — Distinguished Engineer, Yahoo! Vespa — Journey of Vespa from Sparse into Neural Search
We just recorded this episode of Vector Podcast with Jo Kristian Bergum.
I learn a ton by following Jo Kristian’s stream of papers / code / lessons / ideas and thoroughly enjoyed recording this episode with him — grab a cup of your favorite beverage, listen, learn and enjoy!
A few high-level topics from the episode:
- History of Vespa
- Tensor data structure and its use cases
- Multi-stage ranking pipeline
- Game-changing vector search in Vespa
- Approximate vs exact nearest neighbor search tradeoffs
- Misconceptions in neural search
- Multimodal search is where vector search shines
- Power of building fully-fledged demos
- How to combine vector search with sparse search: Reciprocal Rank Fusion
- The question of WHY (my favourite ❤️)
It also turned out, that a few podcasts and public talks I did were relevant to this episode, so you can find them as timed cards in the YouTube version. Hope this helps to learn even more on the related topics of vector search techniques and core ANN algorithms.