Chalk

Our team is brilliant, warm, weird, funny, deeply talented... and growing fast! 💚 Here's a quick look at some of the things Chalk overachievers have accomplished.

A sign of powerful software is when you can replace multiple languages, frameworks, and repositories with one tool applying a clever abstraction, and doing so makes things simple, cheaper, and faster. I can only think of one or two examples. Right now, I'm seeing this with Chalk as it replaces big parts of a pretty serious enterprise ETL pipeline. (And no, I'm not affiliated with Chalk. Their software is just legitimately impressive.)

Q: What's something people are always surprised to learn about you? I have moved 45+ times and lived in 4 countries. 🇺🇸 🇪🇸 🇩🇪 🇨🇷

Q: What made you say yes to Chalk? The product has unlimited potential, and I’m excited to be back in an early-stage environment where I can help build — not just operate.

We're delighted to welcome Caprice Turner to our SF team as Director of Account Management! Caprice is a proven leader in scaling customer success operations at high-growth B2B SaaS companies. We asked her a few questions to get to know her better:

Q: What's the last great book you've read? I just finished Homage to Catalonia by George Orwell. It does a great job of stripping away the mythology of war to give readers an unsentimental view of what it was like to serve on front-lines in a time of incredible social upheaval and consequence

Q: What drew you to Chalk? First, the technology. I think Chalk has a category-defining opportunity ahead of it, and that’s an opportunity I relish. Second, the stage. I’m a builder at heart. Helping define how to deliver value at scale to Chalk customers and, in turn, enable them to deliver unprecedented value to their own customers, is exciting. Third, the people. I do my best work when I’m surrounded by kind, driven people who are smarter than me.

We're thrilled to share that Kyle Koch has joined our LA office as VP of Account Management! Kyle brings over a decade of experience scaling customer success at hypergrowth companies. He's led global strategic customer programs at Databricks and helped scale Sprinklr to a $2B+ valuation. We asked him a few questions to get to know him better:

CHANGELOG: A few updates from the eng team this week 🐴 ➡️ Search your deployment source code directly with Command+K, for faster debugging workflows. ➡️ Offline queries now support numWorkers, so teams can control shard execution at the query level for more predictable backfills and scheduled jobs. ➡️ Combined aggregations are now on by default in the query planner, dramatically improving performance for offline queries over high-cardinality datasets. #february-23-2026" target="_blank" rel="nofollow noopener">docs.chalk.ai/docs/changelog…

Now live: How @Turo built a self-serve ML feature platform for dynamic pricing, and search and ranking with Chalk 🚗 Full customer story here: chalk.ai/customers/turo

Long-awaited team dinner last week with some new faces in SF. Greek food (along with a lot of wine) seemed like the right call!

We're thrilled to welcome Dovie Shalev to our Forward Deployed Engineering team in NY! Dovie joins us from Draper, where he built web-based situational awareness software and worked on system architecture problems. A UMich CS grad with a love for solving proofs, Dovie is also an avid sci-fi and fantasy reader who has recently taken up making kitchen knives.

Most ML teams write the same code twice 🐫 Data scientists prototype in Python. Engineers rewrite for production. By the time it ships: multiple rewrites, multiple systems, nobody has the full picture. This is what MLOps was created to solve. We broke down the challenges + how teams are eliminating rewrites entirely: chalk.ai/for/mlops

📘 Chalk for MLOps – Get full observability with automatic lineage, versioning, and tracing on every query. Deploy in your cloud with enterprise-grade security. chalk.ai/for/mlops

📘 Chalk for AI/ML Engineers – Build LLM applications that combine structured data, embeddings, and vector search. Serve models with sub-5ms feature retrieval. chalk.ai/for/ml-enginee…

We just published four new resources for teams running ML in production, tailored to the unique challenges each role faces:

Ever since we moved to downtown SF, we've been on the hunt for a new bar to play poker, grab wings and hang out. Safe to say we finally found our Friday night spot!