Astronomer, the corporate behind the Apache Airflow-powered information orchestration platform Astro, has secured $93 million in Sequence D funding as enterprises more and more search to operationalize AI initiatives by means of higher administration of their information pipelines.
The funding spherical was led by Bain Capital Ventures, with participation from Salesforce Ventures and current traders together with Perception, Meritech, and Venrock. Bosch Ventures can be in search of to take part within the spherical, reflecting industrial curiosity within the expertise.
In an unique interview with VentureBeat, Astronomer CEO Andy Byron defined that the corporate will use the funding to expedite analysis and improvement efforts and develop its international footprint, significantly in Europe, Australia, and New Zealand.
“For us, that is only a step alongside the best way,” Byron stated. “We wish to construct one thing superior right here. I couldn’t be extra enthusiastic about our enterprise companions, our clients, our product imaginative and prescient, which I believe is tremendous robust in going after collapsing the information ops market.”
How information orchestration turned the hidden key to enterprise AI success
The funding targets what trade analysts have recognized because the “AI implementation hole” — the numerous technical and organizational hurdles that stop corporations from deploying AI at scale. Information orchestration, the method of automating and coordinating advanced information workflows throughout disparate programs, has change into a vital part of profitable AI deployments.
Enrique Salem, Companion at Bain Capital Ventures, defined the crucial challenges going through enterprises in the present day: “Each firm operates a sprawling, fragmented information ecosystem—utilizing a patchworks of instruments, groups, and workflows that wrestle to ship dependable insights, creating operational bottlenecks and limiting agility. On the coronary heart of this complexity is orchestration—the layer that coordinates all these transferring items.”
Salem famous that regardless of its significance, “in the present day’s orchestration panorama is the place cloud infrastructure was 15 years in the past: mission crucial, but fragmented, brittle and infrequently constructed in-house with restricted scalability. Information engineers spend extra time sustaining pipelines than driving innovation. With out strong orchestration, information is unreliable, agility is misplaced, and companies fall behind.”
The corporate’s platform, Astro, is constructed on Apache Airflow, an open-source framework that has seen explosive progress. In response to the corporate’s just lately launched State of Airflow 2025 report, which surveyed over 5,000 information practitioners, Airflow was downloaded greater than 324 million occasions in 2024 alone — greater than all earlier years mixed.
“Airflow has established itself because the confirmed de facto customary for information pipeline orchestration,” Astronomer SVP of Advertising Mark Wheeler defined. “Once we have a look at the aggressive panorama within the orchestration layer, Airflow has clearly emerged as the usual answer for transferring trendy information effectively from supply to vacation spot.”
From invisible plumbing to enterprise AI spine: The evolution of information infrastructure
Astronomer’s progress displays a transformative shift in how enterprises view information orchestration — from hidden backend infrastructure to mission-critical expertise that permits AI initiatives and drives enterprise worth.
“BCV’s perception in Astronomer goes means again. We invested within the firm’s seed spherical in 2019 and have supported the corporate through the years, now culminating in main their Sequence D,” Salem stated. “Past the spectacular progress, Astronomer’s information orchestration has change into much more necessary within the age of AI, which requires scalable orchestration and mannequin deployment automation amidst a ballooning sea of information instruments that don’t speak to one another.”
In response to the corporate’s inside information, 69% of consumers who’ve used its platform for 2 or extra years are utilizing Airflow for AI and machine studying purposes. This adoption fee is considerably larger than the broader Airflow neighborhood, suggesting that Astronomer’s managed service accelerates enterprise AI deployments.
The corporate has seen 150% year-over-year progress in Astro (managed SaaS platform) annual recurring income and boasts a 130% web income retention fee, indicating robust buyer growth.
“Whereas market analysts could also be in search of a transparent winner within the cloud information platforms battle, enterprises have clearly chosen a multi-solution technique—identical to they earlier decided that multi-cloud would far outpace standardization on any single cloud supplier,” Wheeler defined. “Main enterprises refuse to lock right into a single vendor, choosing multi-cloud and numerous information platform approaches to remain agile and benefit from the newest improvements.”
Inside Ford’s large AI operation: How petabytes of weekly information energy next-generation autos
Main enterprises are already leveraging Astronomer’s platform for classy AI use instances that might be difficult to implement with out strong orchestration.
At Ford Motor Firm, Astronomer’s platform powers the corporate’s Superior Driver Help Techniques (ADAS) and its multi-million greenback “Mach1ML” machine studying operations platform.
The automotive big processes a couple of petabyte of information weekly and runs over 300 parallel workflows, balancing CPU- and GPU-intensive duties for AI mannequin improvement throughout a hybrid public/non-public cloud platform. These workflows energy all the pieces from autonomous driving programs to Ford’s specialised FordLLM platform for giant language fashions.
Ford initially constructed its MLOps platform utilizing Kubeflow for orchestration however encountered important challenges, together with a steep studying curve and tight integration with Google Cloud, which restricted flexibility. After transitioning to Airflow for Mach1ML 2.0, Ford experiences dramatically streamlined workflows and seamless integration throughout on-premises, cloud, and hybrid environments.
From AI experiments to manufacturing: How orchestration bridges the implementation divide
A typical problem for enterprises is transferring AI from proof-of-concept to manufacturing. In response to Astronomer’s analysis, organizations that set up robust information orchestration foundations are extra profitable at operationalizing AI.
“As extra enterprises are working ML workflows and real-time AI pipelines, they require scalable orchestration and mannequin deployment automation,” Salem defined. “Astronomer delivers on this in the present day, and because the orchestrator, is the one system that sees all the pieces taking place throughout the stack — when information strikes, when transformations run, when fashions are educated.”
Over 85% of Airflow customers surveyed count on a rise in external-facing or revenue-generating options constructed on Airflow within the subsequent yr, highlighting how information orchestration is more and more powering customer-facing purposes somewhat than simply inside analytics.
This development is obvious throughout industries, from automotive to authorized expertise corporations which might be constructing specialised AI fashions to automate skilled workflows. These organizations are turning to Astronomer to deal with the advanced orchestration challenges that come up when scaling AI programs from prototypes to manufacturing environments serving 1000’s of customers.
Strategic expertise growth: Airflow 3.0 and cloud partnerships place Astronomer for market management
The corporate just lately introduced the final availability of Airflow 3.0, which it describes as “essentially the most important launch in Airflow’s historical past.” The replace introduces a number of transformative capabilities designed particularly for AI workloads, together with the flexibility to run duties “wherever, any time, in any language.”
“Airflow 3.0 lays the inspiration for executing duties on any machine, on-prem or within the cloud, triggered by occasions throughout the information ecosystem,” Byron defined. “It additionally introduces a proof of idea for outlining duties in languages past Python, vastly bettering information crew agility and facilitating migration from legacy programs to Airflow.”
Astronomer has additionally expanded its trade partnerships, just lately attaining the Google Cloud Prepared – BigQuery Designation, making its platform accessible for buy straight from the Google Cloud Market. This permits current Google Cloud clients to expedite their buy of Astro and use their current Google Cloud commit credit.
“We’ve simply signed an superior partnership with IBM,” Byron informed VentureBeat. “They’re placing us into their broader information portfolio of merchandise. And we predict there’s an superior alternative for us, not solely in North America, however internationally, to get a number of momentum with IBM as nicely.”
Unified DataOps: The following evolution in enterprise information administration
Salem believes Astronomer is positioned to redefine enterprise information operations, transferring past orchestration to what the corporate calls “unified DataOps” — a complete method integrating observability, high quality administration, and governance right into a single platform.
“We invested in Astronomer in 2019 with a easy guess: Airflow would change into the usual for information orchestration,” Salem stated. “At this time, it runs at over 80,000 corporations and drives 30 million downloads a month. We backed Astronomer as a result of they’re not solely driving that wave; they’re constructing the enterprise management airplane on prime of it.”
For enterprises struggling to appreciate worth from their AI investments, Astronomer’s progress alerts a vital shift in how information infrastructure is constructed and managed — one the place orchestration serves as the inspiration for the complete information stack.
“As AI raises the stakes for dependable, scalable information infrastructure, we’re doubling down on our funding,” Salem concluded. “Orchestration is simply the beginning. The crew at Astronomer are poised to unify the complete DataOps stack.”
