Chalk

We're thrilled to welcome elvis kahoro to the Chalk SF team! Elvis is joining us as a developer advocate, where he'll focus on expanding Chalk's presence and making life easier, smoother, and more productive for Chalk developers. Welcome to the team Elvis! 👋

San Francisco may have been hit with a tornado warning this weekend, but the Chalk team is here to blow you away with exciting updates! 🌪️ ✨ New features we’ve shipped: 1) Fuzzy text search now available in the source code viewer on the Deployments page—find your code even faster. 2) DatasetRevision objects now support get_metadata and set_metadata methods to manage metadata as dictionaries. Perfect for tracking dataset ingestion and tagging. 3) More array functions added to the chalk.functions library, including array_max, array_min, array_sort, and array_distinct. For more details, check out our full changelog in the link below 👇

⚡ We’ve been busy shipping some exciting new features, so we just published a product update with some highlights of what we've been up to over the past few months: - Enhanced underscore expressions functionality for simplifying feature engineering workflows and boosting performance - Improved dashboards with more metrics and observability for comprehensive insights  - Upgrades to offline queries for smoother workflows - Integration testing support with ChalkClient   For the full update - check out our blog post (linked in the comments)👇 💡 Many more features shipping soon - we always want to hear your thoughts - drop us a line!

Exciting news from Chalk NY HQ! 🏙️ We’re welcoming Daniella Lang to the Chalk family as our first marketing hire. With a track record of building marketing teams from scratch and a knack for product marketing, Dani brings a dynamic energy to our growing team. Welcome Dani – we’re beyond excited to have you with us!

Last week Chalk had its first bi-coastal Holiday Party near our San Francisco HQ 🎅 🥳 ☃️ It was incredible to get the NY and SF teams in one place working together, while capping it off with an evening of great food, drinks, cheer (and some really snazzy outfits). Here's to 2025! ✨ (Not Pictured: Several members of the Chalk GTM team who unfortunately all ate a bad batch of oysters... 🦪)

Recently, an ML leader at one of our partners, Vital.io, summed up his Chalk experience as simply "doing more, with less code." This Thanksgiving, like every year, we couldn't be more thankful for our amazing customers we get to build alongside with everyday. But it's really the moments like these that keep us working tirelessly to simplify ML for developers everywhere. From everyone at Team Chalk, have a great and restful holiday! 🦃 🍁

The atmospheric river flooded San Francisco, so we flooded our codebase with a downpour of new updates. 🌊 ⛈️ 🌊 ⛈️ 🌊 ⛈️   💨 Underscore expressions now support more  chalk.functions for working with arrays and Dataframes, mathematical operations, encoding, formatting datetime, and strings.  😶🌫️ You can now choose whether to cache nulls or default values in the online store with the cache_nulls and cache_defaults parameters. Customers with Redis or DynamoDB online stores can also select to evict null/default feature values for any null/default feature value that would have existed in the online store.  🗺️ You can now define Chalk features as map types, for example user_preferences: dict[str, bool]  🎣 In addition, you can now retrieve Map document types from DynamoDB data sources as either dicts or strings. As always, more detail and much more linked in the full changelog in comments. From the Chalk team, have a wonderful holiday! 🦃

What's the best dashboard you've ever used? We're rebuilding ours and like to learn from the best. Some of our favorites: 1/ Attio / Alexander Christie and team have THE best global search 2/ Chalk / Elliot Marx and co have the best views for the 500+ features we have 3/ Linear / command center + key shortcuts - shoutout Tuomas Artman 4/ New Relic / most customizable and interactive charts 5/ CelerData / it's simple and powerful - no nonsense

In honor of National Apple Cider Day, National Vichyssoise Day, and National Princess Day, pour yourself some hot apple cider 🍎 and a hot bowl of potato soup 🥣, put on your tiaras 👑, and check out what we've been up to at Chalk! ⬛   👶 You can now view the Kubernetes pods created by each deployment in the dashboard along with additional details like the pod states and resources requested by each pod!  🍡 We've added the array_agg function chalk.functions to help you resolve list features with underscore expressions! To see all of the chalk.functions that you can use in underscore expressions for fast feature computation in statically compiled C++, check out our API docs!  🧾 Users can now use the chalk usage commands to view usage information for their projects and environments through the Chalk CLI. As always, the full changelog is linked below or ping us directly if you have any questions 🕶️

While everyone's excited about generative AI (everyone...everywhere..) there are two major, less sexy, challenges that often don't get discussed: Cost at Scale For high-volume applications like recommendation systems (think 300k+ predictions/second), using something like OpenAI's API would cost thousands per second. That's orders of magnitude too expensive for most use cases for most companies. Latency Issues Many applications need responses in ms. Current GenAI APIs take seconds to respond - achingly too slow for many real-time applications like detecting fraud or routing an ambulance. There's real reasons to get excited about GenAI and its application in complex and real-time predictions. The reality however? Traditional ML models still dominate production systems for good reason.