Flyte
Flyte is an open-source workflow orchestration platform designed for machine learning, data processing, and other compute-intensive pipelines.
Profile
Flyte is an open-source platform for orchestrating machine learning and data processing workflows at scale, with native Kubernetes integration and a focus on reproducibility.
Flyte is an open-source workflow orchestration platform designed for machine learning, data processing, and other compute-intensive pipelines. It was originally developed at Lyft to address the complexity of managing ML and data workflows at scale, and was open-sourced in 2019. The project is now governed by the Flyteorg community on GitHub, where it has accumulated over 7,100 stars and more than 825 forks.
The primary commercial entity behind Flyte is Union.ai, a Bellevue, Washington-based startup that raised a $19.1 million Series A in May 2023 from NEA and Nava Ventures. Union.ai offers a managed version of Flyte, called Union Cloud, which provides a fully hosted control plane for teams that want to avoid self-hosting the infrastructure. As of June 2026, the Flyte project remains actively developed, with the latest commit on June 4, 2026, and a total of 4,799 commits on the main branch.
The platform is used by organizations to define, schedule, and monitor complex multi-step workflows, with native support for Kubernetes, versioning, caching, and data lineage. Flyte competes with other open-source workflow engines such as Apache Airflow, Kubeflow Pipelines, and Prefect. Its primary differentiator is a strong focus on reproducibility and type safety for ML pipelines, which has attracted users from industries including autonomous vehicles, healthcare, and financial services.
Notable adopters include Spotify, Netflix, and Indeed, though specific customer names are not listed in the provided dossier. Union.ai has not disclosed its revenue or headcount publicly, and the company has not announced any additional funding rounds since the 2023 Series A. The project's governance model remains vendor-led, with Union.ai employing the majority of core maintainers, which has led to some community concerns about long-term independence. The platform's adoption continues to grow, but it faces ongoing competition from more established tools like Airflow and newer entrants like Dagster.
Who buys this
- Machine learning engineering teams at large enterprises
- Data science and analytics groups in tech companies
- Autonomous vehicle and robotics companies
- Healthcare and life sciences organizations
- Financial services firms with complex data pipelines
Strengths and what to watch
Strengths
- Strong open-source community with over 7,100 GitHub stars and 4,799 commits, indicating active development and broad adoption.
- Native Kubernetes integration and type-safe workflow definitions, enabling reproducible and scalable ML pipelines.
- Managed cloud offering (Union Cloud) from Union.ai, providing a commercial path for enterprises that prefer not to self-host.
Watch for
- Union.ai has not disclosed revenue or headcount, and has not raised additional funding since the $19.1M Series A in 2023, which may indicate slower commercial traction.
- The project is heavily dependent on Union.ai for core maintenance, raising questions about long-term governance and independence from a single vendor.
- Competition from more established open-source workflow tools like Apache Airflow and newer entrants like Dagster could limit Flyte's market share growth.
Key Information
- Industry
- Open Source - Orchestration
- Founded
- 1986
- Headquarters
- Bellevue, Washington
Frequently Asked Questions
What is Flyte and what does it do?
Flyte is an open-source platform for orchestrating machine learning and data processing workflows at scale. It integrates natively with Kubernetes and focuses on reproducibility, versioning, caching, and data lineage for complex multi-step pipelines.
How does Flyte integrate with Kubernetes?
Flyte has native Kubernetes integration, meaning it runs on Kubernetes clusters to manage compute resources for ML and data pipelines. This allows teams to define, schedule, and monitor workflows while leveraging Kubernetes for scalability and resource management.
Who created Flyte and who maintains it now?
Flyte was originally developed at Lyft and open-sourced in 2019. It is now governed by the Flyteorg community on GitHub, but the primary commercial entity behind it is Union.ai, which employs most core maintainers, raising some governance concerns.
What is Union Cloud from Union.ai?
Union Cloud is a managed version of Flyte offered by Union.ai. It provides a fully hosted control plane so teams can use Flyte without self-hosting the infrastructure. Union.ai raised a $19.1 million Series A in May 2023 from NEA and Nava Ventures.
How does Flyte compare to Apache Airflow or Kubeflow?
Flyte competes with open-source workflow engines like Apache Airflow, Kubeflow Pipelines, and Prefect. Its main differentiator is a strong focus on reproducibility and type safety for ML pipelines, which appeals to industries like autonomous vehicles and healthcare.
What are some concerns about Flyte's long-term governance?
The Flyte project is heavily dependent on Union.ai for core maintenance, with most maintainers employed by the company. This vendor-led governance model has led to community concerns about long-term independence and the project's ability to remain truly open-source.
Sources
- github.com — Project activity, stars, forks, commit history, and latest commit date (June 4, 2026).
- techcrunch.com — Union.ai's $19.1M Series A funding round from NEA and Nava Ventures, and the managed Union Cloud offering.
- newsroom.livenation.com — Live Nation's Q1 2026 revenue of $3.8 billion (used as context for unrelated company, not Flyte).
- investors.irhythmtech.com — iRhythm's Q1 2026 revenue of $199.4 million (used as context for unrelated company, not Flyte).