3D data keeps growing. Nothing is built to move it.
CoDim is data infrastructure for the physical world: processing, streaming, analysis, and collaboration for massive engineering data, for engineers and their agents.
Why we started CoDimensional
Teams in simulation, geoscience, robotics, and inspection all hit the same wall. Their data is too large to pass around as files and too spatial for tools built around tables. Nothing on the market moves and processes it, so each team staffs an engineer on a homegrown pipeline that does half the job, then rebuilds it every time the data grows or has to reach someone new.
The wall is getting closer. AI solvers, scanners, and robot fleets now produce 3D data faster than specialists can open it, and agents are starting to read it directly. Engineers will spend less time operating tools and more time asking questions of the data.
The hard part is everything between the data and whoever needs it next.
We are building CoDim as that layer. It processes and analyzes massive 3D data on cloud GPUs and streams it between solvers, agents, ML pipelines, and the tools people use, in any direction. Teams build their products on it through our SDK and API, and their engineers get back to the science.
Leadership
Bane Sullivan CEO · Co-founder
Creator and lead maintainer of PyVista. Geophysicist and software engineer who has spent a decade building 3D tools for engineering and geoscience teams. PyVista users have been asking him to build the rest for years. @banesullivan
What we believe
Data moves in every direction.
Results flow from a solver to an agent, from a scan to a training pipeline, and from the cloud to a browser and back. The infrastructure should carry all of it, wherever it goes next.
People and agents share one source.
Agents will run more of the analysis, and engineers still need to see a result before they trust it. Both should work on the same live data at the same time, with no copies drifting apart.
Every result traces back to its source.
Every view, slice, and number resolves from the source data and the operations that produced it. A decision should never rest on a screenshot that has lost track of where it came from.
The data stays where it lives.
Your buckets, your cloud account. CoDim reads data in place and streams each session only the part it needs, so the whole team can work on datasets far too large to download or copy locally.