turbopuffer is the secret sauce of our gtm knowledge base. such an amazing tool and team

Building a knowledge base and powering all of our agents with it has made them considerably smarter. Getting the right tokens in your context window pays dividends and turbopuffer has made this easy. Cameron Youngblood is working on v2 and expanding scope to include product

i work for a company who's ceo maintains a comprehensive database of 150+ different kinds of teas

OpenClaw + Turbopuffer = perfect recall: strangerloops.com/semantic-memor…

“Don’t worry, there will still be great jobs even when AI automates everything” The jobs:

GitHub status: ❌ Issues ⚠️ Pull Requests ⚠️ API Requests ⌛️ Actions 💔Copilot ⛓️‍💥Webhooks ✅ Sending me irrelevant Slack DMs

we review blog posts at turbopuffer so hard. my git commit history is like "feedback" "more feedback" "small tweaks" "final feedback" "actual final feedback" "nits" "okay i think we're good" "one more"

you can build a highly-available job queue with pretty good throughput in a single file on S3

what if

you’ve heard of “semantic search”, but did you know about “romantic search”? now available on all tpuf regions

I love this

this felt inevitable, and pretty interesting. cloudflare adds markdown, google builds an MCP. early use case I’m thinking about: giving agents control over your website pricing calculator.

sama says he’s a genius but he doesn’t use the oxford comma. not sure what to think

see you next week ;)

tpuf engineers are constantly perfmaxxing

I love Marc - this is such a fun reply

what do you want to ask Simon & Sam? We’re live again next week.

normally I hate tweeting about pricing changes but...

you can judge the quality of an engineering blog by its count of links to academic research papers

Notion scaled AI search by 10x in less than 2 years 🤯 turbopuffer helped them cut costs by 90% along the way