Most tech events optimize for headcount. We're optimizing for depth with our data λ code unconference.

I'm personally reviewing every RSVP to build a room of practitioners ready to:
💡 Share real technical challenges
💡 Propose sessions on problems they're solving
💡 Engage thoughtfully in discussions

Data Debug SF round two exceeded all expectations. Only our second event and the conversations were awesome! I also loved seeing familiar faces plus new people discovering what we're building.

Easiest way to get a stakeholder to ghost you, ask them:

"Can someone review this data change?"

Stakeholders will ignore abstract requests. Show them how their dashboard changes or give concrete examples about why it matters to them.

Send screenshots + the "so what" to get their attention.

Human-in-the-loop reviews + automated tests are the best way to avoid having data nightmares & reverting code.

Anyone else felt this gif in their bones?

The data-analyst-in-a-box race is heating up, and Snowflake entered last week with their own AI agent play.

I've seen four waves of this play out so far. We're currently in Wave 4.

Today I'm wandering around the Databricks Data + AI Summit with a few Recce conference survival kits. They're packed with conference essentials I always carry in my backpack, but now in one goodie bag with Recce stickers :)

If you see me, come up and say hey to get one!

Snowflake dropped what they are calling AISQL.

AI_CLASSIFY immediately jumps out at me. I've spent so much time manually crafting CASE/WHEN statements that IFF(command works, big time saver, more noise).

Yes, I did make a SQL pun of my favorite SnowSQL CASE/WHEN replacement IFF().

Just wrapped up the week in Boston at the ODSC East conference. Unsurprisingly, AI and AI-related topics were probably 95% of the talks. What was surprising, a lot of attendees were from non-tech companies. Even older institutions and companies in non-tech industries want their DS to use AI.