Will Kenney retweeted
we're changing tpuf's storage architecture so we can keep pushing the search frontier
the new engine (v3) now passes CI, but we have to make it fast before we ship it to prod
we're grinding query plans in public until v3 wins on everything
follow along: turbopuffer.com/v3
Will Kenney retweeted
we're puffin
we partnered with @appliedcompute to post-train a small model for large-scale code search over precomputed indexes
at 300 repos, this is ~3x faster than using filesystem + grep, and reduces the marginal cost of a search by up to 100x vs frontier models
turbopuffer.com/blog/large-s…
Will Kenney retweeted
A 35B open-weight model trained to search a precomputed index answers repo search questions at 100x lower cost than a frontier model.
We partnered with @turbopuffer to train Qwen3.6-35B-A3B to find code across ~9,000 repositories. It tops the needle-in-a-haystack task outright at 2-10x lower latency.
Will Kenney retweeted
turbopuffer is one of my favourite companies! Not only is their product insanely performant , their merch game is lowk unrivaled
fal just made it possible to generate video faster than you can watch it
turbopuffer makes it possible to search every video, image, and 3D model you generate on @fal
→ 300M+ assets
→ 225k+ namespaces
→ 17ms p50 hybrid latency
customer log: [tpuf.link/fal]
Will Kenney retweeted
just got back from a weekend of "summer camp" in Algonquin with the tpuf crew + families
the wifi & weather was terrible but the vibes immaculate
my daughter threw a "STAY STAY STAY CAMP" themed tantrum at the end
Will Kenney retweeted
How can you a product that is 10x cheaper or 10x faster than what is already out there? @Sirupsen (cofounder and CEO at turbopuffer) used "napkin math" - doing quick calculations to get rough answers - to get here.
Full interview from AIE: youtu.be/h8TBLKyo7Rs
Deepdive in The Pragmatic Engineer: newsletter.pragmaticengineer…
Will Kenney retweeted
we are releasing our first open source project! puffgres is a sync service between Postgres and @turbopuffer. instead of two DB calls everywhere, you write ‘dumb’ JS transforms (i.e. tokenization, embedding) and the core handles logical replication, retries, batching, etc
Will Kenney retweeted
our team is remote but our roots are Canadian. when it came time for turbopuffer's first feature, Canada's The Globe just felt right.
the piece is a window into our founding story, business, and customers.
we're a lean team of 38 + thousands of happy customers. puff with us!
Globe&Mail wrote a generous piece about turbopuffer theglobeandmail.com/business…
Will Kenney retweeted
Worklife Update: I’m spending the summer with @turbopuffer 🐡☀️
This hardcore whimsical team (of only 38!) is doing something special and their product is exploding.
The plan is to get even more people puffing and write about the experience.
What do you want to learn about this mostly-bootstrapped pretty-Canadian AI darling making search cool again?
Will Kenney retweeted
this spring, we've been having a lot of fun making some puffy inspired pieces for our customers who've been puffin' along with us
drop #002 now live on turbopuffer.supply 🧱
Will Kenney retweeted
my second favorite part was covering @willkenney in snow for the opening shot
my favorite part was witnessing Zola and Tarun seeing snow fall for the first time 🥹
⛷️ turbopuffer summit #004 🏂
our remote team spent a whole week together in banff national park. we brought a camcorder because life is short and why not:
Will Kenney retweeted
we couldn’t afford the sphere so we got the next best thing
costs less, hits harder
Will Kenney retweeted
Cursor has a huge indexing pipeline. big repos were taking too long to index for semantic search
they cut index time for massive repos from hours to seconds using tpuf's `copy_from_namespace` aided by some clever hashing
btw, we're working on an api to make this even faster
Semantic search significantly improves coding agent performance.
For very large codebases, Cursor's indexing process is now several orders of magnitude faster.
cursor.com/blog/secure-codeb…
Will Kenney retweeted
An incredible year for turbopuffer is closing. We now manage trillions of vectors and tens of petabytes. We’ve puffed up revenue 10x and headcount 5x.
We’ve helped Anthropic, Cursor, Notion and others connect every petabyte to AI to ship the most ambitious version of their product.
This week we doubled down with Lachy Groom and added Thrive Capital to the team. They have both proven themselves repeatedly to us over the past 2 years.
In 2025, we’ve proved the puffer architecture is capable of state-of-the-art vector and text search. Next year, we’ll continue to optimize the core and widen our surface area, to make it effortless to search every byte.
Will Kenney retweeted
.@emmetkenney caps off the drive with 3⃣ from 23 out.