Get the latest tech news
Turbopuffer: Fast search on object storage
turbopuffer is a vector database built on top of object storage, which means 10x-100x cheaper, usage-based pricing, and massive scalability
Let's take a step back to see how search engines fit into the broader infrastructure stack and how their requirements differ from other types of databases. Justin (CTO) and I learned the hard way during our 5+ years on the last-resort pager at Shopify that the fewer stateful dependencies, the more nines of uptime. If you want our HA number, I’ll just send you the current kubectl get pods | wc -l. Additionally, because much of our compute runs on spot nodes, we constantly exercise this failure path.
Or read this on Hacker News