Kunal Batra

The @datadoghq DASH fireside chat lineup looks amazing. @oliveur + @ClemDelangue Alexis Lê-Quôc + @thsottiaux @atalwalkar + @_sholtodouglas Yanbing Li + @tomocchino I absolutely can't wait. See you in June! dash.datadoghq.com

👀

Heading to @iclr_conf next week? Datadog AI Research is hosting an event along with the ICLR SPOT workshop organizers. Join us to meet our team and learn about our work on observability world models + agentic post-training!

Been working on this for a bit, stoked to announce a community port of the AWS Strands Agents framework to Swift. It can run natively on device and you can build those fun OpenClaw-style computer use examples. It can run inference locally via MLX or through Bedrock. In the video I connected an agent to a Mac MCP server and asked it to open Word and write a sentence. It also sends OTel telemetry to your tool of choice. In the video you can see the telemetry from the agent being shown in Datadog's LLM Observability. communitystrands.com/swift Coming soon: Interacting with the agents through voice either AWS Nova Sonic or locally with @Prince_Canuma's mlx-audio-swift package. cc: @awsdevelopers, @awnihannun , @shshnkp

Introducing MLX-Swift-TS github.com/kunal732/MLX-S… An SDK for running time series foundation models fully on-device on Apple Silicon. When I joined @datadoghq , I was introduced to Toto, our time series foundation model, and got excited about zero-shot forecasting across different domains. While building a health copilot app, I realized there wasn’t a simple way to run models like these locally on device. So I built one. MLX-Swift-TS exposes a common TimeSeriesForecaster interface for loading and running multiple time series architectures directly in Swift using MLX. No server required. The attached video shows on-device forecasting running inside a native Swift app. Huge thanks to @awnihannun and the MLX team for building MLX and its Swift API, @Prince_Canuma for inspiration on MLX SDK patterns, and @atalwalkar and the Datadog team for Toto.

@mgonto Do you have any frameworks you prefer for constructing/running these ?

Every time I explain growth to people, I explain them that you're betting on an experiment to have an increase. Because of that, every quarter you have big bets, medium bets, and small bets. Big bets are big swings that take you a lot of time and might get a huge uplift, but you don't know if they're going to work out or not. Because of that, you do small bets like changing small things on the website or medium bets which are more likely to work but are not a step change until you can get one of those check step changes to work. Growth is betting

@fullstackpython Ha fair - “almost” does a lot of work there 😉 Always appreciate a good meme though.

@kunal732 I said “almost” 😉

Almost as if it was designed that way 🤔

Always enjoy chatting with this crew

@goosewin @RhysSullivan Ha

@RhysSullivan @kunal732 i told you

bros got a datadog

@shshnkp @awnihannun @angeloskath This is great!

Exciting MLX updates at #WWDC25 new webpage and 2 sessions for Python and Swift devs! New MLX webpage! mlx-framework.org Getting started with MLX by @awnihannun developer.apple.com/videos/play/ww… Explore LLMs on Apple silicon by @angeloskath developer.apple.com/videos/play/ww…

@crtr0 same ✉️

No question ☎️

@Prince_Canuma Mlx evangelist !!

🔥

@amit Nice! I see you got the noise canceling ones

New pods 😊

@Prince_Canuma Do you have or know any resources to help convert times series foundation models like google's timesfm to MLX ? From my understanding these use patch embeddings for inputs in a similar way to how image inputs are encoded for vlms .

@sach1n @Prince_Canuma @awnihannun Thank you! Have been spinning my wheels the past couple of days to get 7B-Instruct to work but keep running into issues

Just ported Qwen2.5-VL-3B-Instruct to MLX Swift. Many thanks to @Prince_Canuma for the Python version from which this is based upon. @awnihannun I’ll do further testing and send out a PR later this week.

Great work by @kunal732 to help run locally using MLX the @arcee_ai Virtuoso Lite 10B model distilled from DeepSeek V3.

@mkurman88 @reach_vb This is great! Thank you!

@reach_vb huggingface.co/mkurman/Qwen2.…

Absolutely love how cute DeepSeek distilled fine-tunes are! ❤️ "I think the first option is best"

@vatsal_manot cc: @shshnkp - not sure if there is an internal logging system for feedback like we had at AWS but for developers. But something to note.

Apple’s documentation is often sparse, non-visual and rarely do they ever disclose the list of confirmed bugs by OS version. And that’s for the documentation that exists _at all_. Community sources are often the only source for any real-world sample code for new APIs.

@vatsal_manot It's what makes MLX surprising. Just doesn't feel like its from apple.

@kunal732 I’ve been developing on Apple platforms since 2011 and while these things are never uniform (for example, the MLX team is very proactive at resolving missing documentation), it’s generally been in steep decline on the whole.