marekgalovic

6 ms·

1. 1.▲
   by marekgalovic 1y ago

   We do a projection of the original vectors so that it matches one of our optimized kernel. This generally gives us better recall vs. simple padding since all bits are utilized.
2. 2.▲
   by marekgalovic 2y ago

   We actually tried to extend DataFusion at first but ultimately decided against it since we can get most of the value by using Arrow and its compute kernels directly. DataFusion also executes filters in a way that rebuilds the underlying arr
3. 3.▲
   by marekgalovic 2y ago

   Founder of TopK here. There are legit use cases for vector-based retrieval (e.g. semantic search, recommendations, multi-modal search, etc.) but that only requires supporting vectors as a data type, not building the whole database around ve
4. 4.▲
   AnnDB: A Distributed Approximate Nearest Neighbors Database(anndb.com)

   7 points by marekgalovic 6y ago|0 comments
5. 5.▲
   Take me as an intern during the summer 2016(intern.marekgalovic.com)

   3 points by marekgalovic 11y ago|0 comments