
Live
Built withC++17AVX-512pybind11Python
Vector Vault DB
A vector database written from scratch in C++17
Real-time recommendations on streaming features

Overview
Two-stage retrieval and ranking over features that update as users act, rather than as a nightly batch. Kafka carries the feature pipeline, FAISS handles candidate retrieval, and the whole path is instrumented.
Two-stage candidate retrieval then ranking
Streaming feature pipelines over Kafka
FAISS vector search for candidate generation
Online metrics on serving latency and quality
Keep exploring

A vector database written from scratch in C++17

MLOps control plane that closes the loop
Let's talk
Tell us what you're building. We'll help you scope it, build it, and ship it.