Ben Hudson

Product lead @atlanhq. Previously @dremio, @IBM, @wesleyan_u, @wes_tennis.

[Ben Hudson @benhuds]

Full DML on the data lake feat. @dremio x @ApacheIceberg. Exciting stuff! https://t.co/talmFcofE2

[Ben Hudson @benhuds]

@rachitcodes @PrathmeshPatel_ @rishigb and I have already posted some thoughts on this but specifically for MCP servers 🤠

Part 1: https://t.co/hlYLPXTrFQ

Part 2: https://t.co/pWRwWz4PKI

[Atlan @AtlanHQ]

We’re excited to launch Atlan Trident 🔱

The shared context workspace for humans and agents.

One place for every skill, agent, metric and ontology.
Shared across your team, in the tools they already use.
Better with every run.

[Ben Hudson @benhuds]

@AlexReibman @petergyang's /no-ai-slop skill is a good starting point https://t.co/uOn5w57vpI

[Ben Hudson @benhuds]

@surendranb @surendranb please do contribute a UI! Looks like there's already an issue on the repo mentioning the same https://t.co/hPSI4PJnx3

[Atlan @AtlanHQ]

Today at Context Conference, we open sourced AI registry, our open standard for agent registries.

Ben Hudson (@benhuds), Atlan: "Context needs an open ecosystem. The world we live in today is multi harness, multiplayer, multi model."

Build your own open registry, share feedback, and contribute:
https://t.co/BXOOGdUAfz

[Rishi Gaurav Bhatnagar @rishigb]

On Oct 1 in San Francisco, 5 builders showed us how they're building multiplayer AI: their thesis, their approach, and where they disagree.

Chat or canvas? Who owns quality? What gets shared?

Here's everything from the evening 👇 https://t.co/Qgikv27zfK

[Ben Hudson @benhuds]

@_manaschoudhary It may sound simple but in my experience the best way is to just do the repetitions – whether it's an internal presentation, community talk, or just your regular Slack comms

[Ben Hudson @benhuds]

@johnroodepic @rishigb Yep, tool selection definitely shows up as a failure mode!

[Ben Hudson @benhuds]

@Dipankartnt Brickworks? Looks great!

[Ben Hudson @benhuds]

@alpha_convert @hackedy @wilbowma I remember getting that email! Had to search my inbox to jog my memory but glad the book turned up on their doorstep in the end :')

[Ben Hudson @benhuds]

@mallik70 @Miki__Yoshi @AMdatalakehouse Thanks @mallik70! Interop is a big focus. Aside from Arrow (now a standard in-memory format), another area of interop is quickly/easily retrieving data between systems (DBs, BI tools, etc) via Arrow Flight/Flight SQL. Might be of interest https://t.co/SLpdkcFBX0

[Ben Hudson @benhuds]

@mallik70 @Miki__Yoshi Happy to chat further but hope this provides a bit of context for you @mallik70!

[Ben Hudson @benhuds]

@mallik70 @Miki__Yoshi Openness also holds from the DW vs DL perspective. A little more straightforward since once you load data into your DW you're locked into that DW's functionality, etc. And speaking of lock-in/out, DW migrations are a different beast altogether 😅

[Ben Hudson @benhuds]

@mallik70 @Miki__Yoshi @mallik70 wrt table format contributors, Dremio is just one of many contributors to the Iceberg project (looks like top 3 are Netflix/Apple/AWS). You can see a full list on this blog by @AMdatalakehouse here! https://t.co/xhFpPBnBbB

[Ben Hudson @benhuds]

@mallik70 @Miki__Yoshi @mallik70 👋 Dremio currently supports both Iceberg and Delta Lake table formats — we have customers using both. Arrow is just the in-memory format (we read data into memory as Arrow regardless of file/table format).

[Ben Hudson @benhuds]

@Stephenpedj @__AlexMonahan__ @simonw @bernhardsson Glad to add a bit of context @Stephenpedj! We do have a free SaaS version if you'd like to give it a test drive https://t.co/Zatz2HuAFJ (don't need to know Docker 😬) in any case, please let us know if we can help in any way (get started/use case/etc) 🙌