dbt Labs

350+ packages. 4,200+ versions. One engine to support them all. Getting users onto the dbt Fusion engine is one thing. Getting the entire package ecosystem ready for them is a whole different challenge. Chaya Carey has been six months deep in that work. Anders Swanson sat down with Chaya to talk through what that looks like. They cover: • Why Jinja flexibility is both a superpower and a migration challenge • How Package Hub evolved to surface Fusion compatibility info • What dbt autofix handles automatically so you don't have to • How the team went from manual package reviews to automated Fusion parsing at scale

Agents will be the primary consumers of analytic data within 12 months. I realize that's a big claim. Here's why I believe it. Over the past decade, we invested massively in data infrastructure. It worked — the bottleneck moved. But we just revealed a new one. We never actually fixed *analysis*. Analysis, mostly, is just *thinking*, and we hadn't solved thinking yet. Well, that bottleneck just vanished. Meta went from weekend prototype to a company-wide analytics agent used by thousands — in six months. OpenAI built their own. Ramp built Ramp Research. The dbt MCP server has grown 50% month-over-month since launch. These aren't experiments. They're in production. The analyst role isn't disappearing. But it is fundamentally changing. The analysts who thrive will be building and operating agentic systems, not shipping dashboards. How quickly do you think your org will cross that threshold? I write about this every week in the Analytics Engineering Roundup — link in my profile if you want to subscribe. #AnalyticsEngineering #DataAnalytics #AIAgents #dbt

Who wants to learn how to optimize their compile times in dbt? How many of you have waited 10+ minutes for a CI job just to start running a small state:modified` subset? Well I have 2 recs: 1. Adopt Fusion if you haven't already. Parse times are way faster, same with compile times. 2. Listen to Anders Swanson interview my partner and engineering lead for Fusion, Alexander Bogdanowicz. He'll tell you about how certain patterns in Jinja can ruin performance of compile and cause all commands that depend on it (run, build, etc.) to take unnecessarily long. https://lnkd.in/gXwPkngm

The analytics world is shifting faster than most people realize. This week's issue of The Analytics Engineering Roundup features Tristan Handy's take on where analytics is heading and what to do about it https://lnkd.in/gwMhnXAg

AI-assisted coding is now a top priority for 72% of data teams. AI-assisted pipeline management? Only 24%. Teams are moving fast. The governance layer isn't keeping up, and that's where trusted data lives. The 2026 State of Analytics Engineering report breaks it down https://lnkd.in/gWJsRtTE

dbt Core just hit 1 billion downloads, 10 years after it shipped its first line of code 👏 The past decade has had literal exponential growth (R²=0.998). At this rate, we'll hit 2 billion downloads in about 7 months.

See you at #GoogleCloudNext next week 👋 dbt connects directly with Google BigQuery, AlloyDB, and BigLake, adding testing, documentation, and version control so your pipelines become systems you can trust. The dbt Semantic Layer gives everyone a single source of truth for metrics. No more debating definitions across dashboards and teams. Find us at Booth 6606 to see it all live alongside AI, including our VS Code extension integrated with Google's Antigravity IDE. Book a 1:1, enter our raffle for Apple AirPods Max, or join us April 22 at Eyecandy Lounge (Mandalay Bay) for our afterparty with Fivetran and 66degrees. 🎲 Read the blog for all the details https://lnkd.in/g7YQgSGc

Most financial market infrastructures have the data. The hard part is making it consistent, trustworthy, and ready for AI. Jamie Nemeroff, Andrea DeSosa, and Michael Weiss break down how Nasdaq built a governed intelligence layer across trading and post-trade workflows, using dbt Labs and Databricks, and how they're scaling it across markets without sacrificing data contracts or lineage. Global friendly sessions April 28 and 29. Save your seat https://lnkd.in/g5TytEYf

A powerful combo for modern analytics teams. 🤜🤛 Discover how dbt + Microsoft Fabric integrate, see real‑world engineering patterns, and get practical guidance for any skill level. Register today: https://msft.it/6046QCDNe dbt Labs