The combination gives engineering teams one place to automate their most important work, from data pipelines to agentic workflows
Prefect, the maker of critical AI and data automation software, announced that it has agreed to acquire Dagster Labs, creators of the widely popular Dagster orchestrator. The transaction brings together the two most widely adopted successors to Apache Airflow, creating a combined suite of products that serves thousands of data teams running production workloads across data pipelines, ML operations, and AI agent infrastructure.
Also Read: AiThority Interview with Matej Bukovinski, Chief Technology Officer at Nutrient
Prefect, the maker of critical AI and data automation software, announced that it has agreed to acquire Dagster Labs, creators of the widely popular Dagster orchestrator.
Engineers have always demanded control over the systems they automate. For data teams, that has traditionally meant pipelines; now, increasingly, it means agentic workflows. Prefect runs the work and governs how agents reach tools and data through FastMCP. Adding Dagster’s declarative model for defining outcomes gives the combined company all three parts of the modern automation problem: what work should produce, how it runs, and how agents are governed.
That asset-based model is purpose-built to describe what data and results should exist and to confirm that the work delivered them, which is exactly what defining automated outcomes requires. “For years, Prefect and Dagster raised the bar for each other, and as a result, we built two of the most popular tools in the data ecosystem,” said Jeremiah Lowin, founder and CEO of Prefect. “Together, we’re consolidating the category around our shared vision while also building the best foundation for the next generation of AI automation.”
FastMCP holds a position in the agent ecosystem that no other orchestration company occupies. It has been downloaded more than 92 million times in the last month and has earned over 26,000 GitHub stars, making it the default way to connect AI agents to external tools and data and the basis of most MCP servers in production. Prefect shipped it the weekend after Anthropic announced the Model Context Protocol, built to make the new protocol usable for everyday developers, and it worked well enough that Anthropic adopted it as the official MCP SDK. Prefect continues to work closely with Anthropic’s MCP team, consulting regularly on the Python SDK and its developer experience.
“Prefect and Dagster are the two orchestrators the market turned to as it moved past Airflow, and putting them under one roof creates the clear leader in next-generation orchestration,” said Pete Hunt, CEO of Dagster. “That is what got me excited about this. Prefect is one of the only companies that could credibly keep building Dagster and supporting Dagster customers for the long term, and that mattered more to us than anything.”
What this means for current users
Dagster founder Nick Schrock and CEO Pete Hunt will serve as strategic advisors to Prefect and will remain active in the open-source community.
Both products receive long-term support and continued investment. Both Dagster and Prefect continue as independently supported products, and the commitments are specific:
- Dagster and Dagster+ keep their names, open-source license, and roadmap, maintained and advanced by many of the same engineers and go-to-market people who support them today.
- The features and support that teams rely on today carry over to both products.
- Pricing for both Prefect Cloud and Dagster+ is unchanged. Any future pricing decisions will be shared clearly and with plenty of notice.
- Open-source Prefect and open-source Dagster both continue to receive maintenance releases, new features, and security patches.
- Teams running either product in production can stay exactly where they are, for as long as they like. Any team that wants both can use them together.
“As WHOOP builds more AI-driven experiences for our members and across our internal platforms, we need infrastructure that can govern automated decisions with real confidence,” said Carlos Peralta Head of Data Platforms and ML Ops at WHOOP. “Prefect’s move to steward Dagster alongside FastMCP gives us the right tool for each job, with asset-aware orchestration for our data pipelines, flexible Python-native execution for dynamic workflows, and a modern protocol layer for connecting AI agents to the systems they depend on. That’s the kind of clarity engineering teams actually want.”
Where the combined company will invest
The workloads hitting production orchestration have changed, and the combined engineering team will concentrate on three areas.
A full-stack approach to agentic automation
Prefect brings runtime execution, FastMCP brings governed agent access, and Dagster brings declarative outcomes. Together, they let a team automate an agentic workflow the same way they have always automated a pipeline: defining what it should produce, running it, and keeping it under control. This capability lets teams treat agentic work as a first-class part of their systems.
Reliability moves upstream, into the layer that runs the work
Most reliability tooling reports failures only after they have already occurred. Dagster’s materialization tracking and freshness policies catch data problems at the point of computation, and Prefect’s event-driven execution reacts the moment state changes. Together, they move reliability upstream so teams catch problems while they are still cheap to fix.
Flexible execution meets declarative structure
Prefect and FastMCP were built to be as flexible as possible, capturing code as it is written. Dagster brings the complement: a declarative, compile-time way to describe what work should produce. Pairing Prefect’s flexibility with Dagster’s declarative model gives engineers both a natural way to write their code and a rigorous way to reason about its outcomes.
A profitable foundation
Prefect has operated profitably for the past year. For any team deciding whether to build on this software, that is the reassurance that matters: these products have a home funded by their own revenue, able to support and grow them through any market cycle. For the Dagster community, it means the project’s future rests on a self-sustaining business.
Also Read: AI systems – Interoperable AI systems: Connecting models across platforms
[To share your insights with us, please write to psen@itechseries.com]
Comments are closed.