When Ivan and I started Iterative Inc., the company behind our open-source project DVC, our mission was simple: bring version control discipline to data – the same way Git transformed how we manage code.
Back then, it was chaos. Data scientists were juggling folders, naming files like data_final_v3_really_final.zip, and struggling to reproduce experiments. We wanted to fix that, to make ML workflows reproducible, collaborative, and shareable.
DVC grew into something far bigger than we imagined. It became part of how data scientists think about their work. The community took it from an idea to a global standard. You built tutorials, created tools around it, and pushed its boundaries in ways that constantly surprised us.
That open-source creativity has always been DVC’s real engine.
Different Paths, Same Mission
As a former researcher and data scientist, I’ve always been drawn to the higher layers of the stack – the places where data turns into insight and intelligence. DVC, on the other hand, has naturally evolved deeper into infrastructure.
Over the years, as Iterative expanded into new frontiers like DataChain, our focus evolved toward solving new challenges in the multimodal AI world.
At the same time, lakeFS was growing in the direction of infrastructure, tackling version control for massive enterprise data lakes.
From the first time I spoke with Einat, Oz, and the lakeFS team, we immediately felt a meeting of the minds, with a lot of shared experience, despite coming from different directions. It was obvious: we were solving different parts of the same problem.
Einat said that DVC and lakeFS had “tons of synergy” – and then she smiled and added, “I can be the best mom for DVC.” That line stuck with me. It resonated not just on a technical level, but on a personal one – because DVC has always felt like our child. I couldn’t imagine a better home, or a better “parent,” for it than lakeFS.
- DVC helps individual data scientists and small teams move fast, experiment, version, and iterate – leveraging familiar GitHub and GitLab infrastructure to make data versioning feel natural and lightweight.
- lakeFS brings those same principles to petabyte-scale systems, offering its own SaaS platform optimized for scale, reliability, and governance enterprises need.
Two perspectives. One philosophy. It just made sense that these tools, and communities, should eventually meet.
Why This Transition Feels Right
This isn’t about letting go. It’s about amplifying what DVC stands for.
The lakeFS team shares our obsession with reproducibility, collaboration, and open source. They’ve built a platform that extends DVC’s ideas to a new level of scale – and they deeply respect the community that made DVC possible.
By bringing DVC and lakeFS together, we’re creating a continuum for every data team: from the first ML experiment on your laptop → to production pipelines managing petabytes in the cloud.
No lock-ins. It is a natural evolution from small data to big data, grounded in the same engineering principles.
Gratitude and Looking Ahead
To everyone who contributed to DVC, thank you. You’ve made open-source MLOps what it is today. Every pull request, issue, and blog post shaped a tool that helped thousands of people work smarter with data.
To Einat and the lakeFS team: thank you for carrying this vision forward and for giving DVC the perfect new home. I’m excited to see what happens when these two worlds combine: the creativity of individual data scientists and the scale of enterprise infrastructure.
As for me, I’m excited to explore what’s next with DataChain Inc. – a company inspired by what we learned, but focused on a different mission: advancing data processing and analytics in the multimodal AI world.
DVC was built on a simple belief: data deserves the same care as code. That belief lives on – stronger than ever and on an even bigger stage. Thanks to the Community that made it real.
Here’s to the next chapter! 🚀
– Dmitry
Join our Webinar: A New Chapter for DVC: Passing the Torch to lakeFS
I’ll be joining Oz Katz, CTO at lakeFS for a webinar on December 3, 2025 to discuss the transition. We look forward to seeing you there!
Questions?
For any questions you have on the transition, please see our FAQs blog post.
📰 Join our Newsletter to stay up to date with news and contributions from the Community!