


excel user & data enthusiast. living the dream at MotherDuck


We are pleased to announce: As of today, MotherDuck runs on Microsoft Azure! Our platform weaves together all the features that are needed to deliver data products, all in one place. If your company is an Azure shop, you can now run MotherDuck without bringing in another cloud.

Jsheets fixes this
Slop Join: joining data from multiple MCPs inside an agent context window. I actually see this a lot in GTM teams when they’re pulling marketing data together. We need a better approach for big data through MCP.

tried to go pure claude (docs, sheets, dashboard, artifacts) last couple days and i regret it. complete waste of time but useful for one-offs

sir, they're doing tricks on it
Researchers have used reinforcement learning to teach a 5-tonne John Deere skid-steer loader to perform complex acrobatic manoeuvres autonomously. Trained in simulation and transferred to a real machine, the system can balance on two wheels, navigate rough terrain, spin, and control its bucket before landing safely. The research demonstrates the potential of AI-driven automation for heavy construction and industrial machinery. Video Credit: @leggedrobotics / @ETH Paper: wevlv.co/4kltYys #engineering #technology

its get that bread week
What happens when software engineers and data teams meet in San Francisco to build their own bakery with an agent? The answer is.. we'll ducking see. Space is limited, join us in December if you're around: luma.com/bakeoff

live with fearless leader Jordan Tigani! join us to hear about where things are going with data + AI
Everyone gets a data agent: What's coming in MotherDuck from our CEO x.com/i/broadcasts/2brOBjEnL…

I've spent years herding engineering teams to maintain backlogs. A backlog is inventory, and inventory is expensive. Building with AI agents, I've mostly stopped using one. Some thoughts on how agents are changing the way we decide what to build. Link below.

Your 30s is this insane decade where you enter it a youth and leave it an elder. When you’re 30 you go to the doctor and be like hey doc I’m living on McDoubles and McCallan 12,and they’re like Ah to be young! When you’re 39 you’re like hey doc my tummy hurts sometimes and they’re like we need to get some tests, and btw if you eat another grain of salt you’re going to die. And out of nowhere one day you wake up in your late 30s and every cop and firefighter and nurse is a teenager. Somewhere in your early 30s cops start calling you Sir and it’s like ok enough with the joke. And then in your late 30s some cop who just got out of middle school calls you Sir and you’re like good, that’s the respect I deserve out here. In retrospect your 20s is like a remarkably stable decade. You enter a young idiot had leave a young idiot. Your 30s is a wild ass ride especially when they throw a pandemic in the middle of it.

Data Intelligence Engineerという新しい役割についてのポスト。 ・データチームの仕事は、ビジネスが良い意思決定をできるようにすること ・これまでは問いが答えになるまで、マーケ担当→アナリスト→データエンジニアとチケットのリレーが続き、届く頃には聞いた理由を忘れていた ・2025年12月より前は、LLMはSQLが下手で、データ分析はAIの影響を受けにくかった ・しかしモデルのSQLが良くなり、業務ロジックは文脈があれば推測でき、エージェントがマルチターンになって試行錯誤できるようになった ・SQLを書く、ダッシュボードを作る、dbtパイプラインを組むといった作業は、専門知識がなくてもできるようになっていく ・社内のデータエンジニアは、Airflowのバージョン移行を最小限で週末に終えた。以前なら1週間かかったはず ・それでも、正しいとサインオフする責任は人間に残る ・データエンジニアリングは説明責任の機能 ・モデルはSQLは知っていても、会社が売上をどう数えるか、似たテーブルのどれが本物かは知らない ・コンテキストはエージェント向けのマニュアルで、まず綺麗なデータモデルを作るほうが良い ・Data Intelligence Engineerは、エージェントを作って取り込みや変換を自動化し、組織の知識を共有コンテキストに書き込む ・evalsで、答えの分かっている質問に正しく答えるかを確かめ、壊れたことにも気づけるようにする ・ゴールはセルフサービス ・誰でもエージェントでビジネスの問いに正しく答えられるようにすること motherduck.com/blog/the-data…

Mom come pick me up they’re arguing about semantic layer vs ontology in the group chat




ok like this is actually...art? this is really good
Playing with Fable 5.2 and @jamesonhodge’s Chimera vocal synth. It's always nice to hear Bach in a new adaptation.

M364 amiright