About
We build backend infrastructure for robotics deployments.
Most robotics teams do not stall on model ideas first. They stall in the operational loop: collecting trustworthy data, reproducing failures, and shipping fixes without guesswork.
Thesis
Thesis: the deployment gap is operational
Research demos keep improving. But real deployments still fail on the long tail, because the learning loop breaks under real constraints.
When a demo looks good, the production questions are unglamorous: what happens when lighting changes, a camera shifts, the network drops, or an operator makes a mistake? Did you capture the right data, and did it arrive intact? When you have weeks of recordings, can you aggregate them into something you can browse, search, and slice again later?
Most teams end up stitching together a fragmented stack: loggers, object storage, ad-hoc scripts, a visualization tool, a training pipeline, and a spreadsheet of “what happened.” It works for a while. Then the fleet grows, the data grows, and nobody can keep the system coherent. People argue about what was collected, what was dropped, and which version of a dataset a result came from.
We think the missing layer is a deployment backend: a data + ops system designed for robotics semantics and edge reality. Its foundations are reliable collection, aggregation you can trust (sessions, manifests, and metadata), and fast retrieval.
Collect → Curate → Learn → Deploy → Monitor → Improve → repeat.
Compounding isn’t automatic. If failures force human intervention, intervention drives cost. Cost limits deployment scale. Limited scale means less real data. Less real data keeps the long tail unsolved.
The way out is to make the loop operable: capture data reliably in the real world, aggregate it into coherent recordings and datasets, and ship improvements safely.
That’s what we’re building.
Scope
What we build
DataCore
In pilotThe data and retrieval backbone: reliable edge-to-cloud capture, robotics-native indexing, and synchronized slices for debugging and training.
Programs (with partners)
In progressPartner deployments that test reliability and retrieval performance in production environments. These are standardized product engagements, not custom services.
Deployments
PlannedOver time, we expect to run selected deployments on the same stack to harden it under real constraints while staying focused on infrastructure.
Operating model
Principles
Build for messy reality
If it only works on a good network, it won’t work where robots are.
Make every fleet hour improve the next release
Data should compound. Debugging should be fast. Datasets should be reproducible.
Design for autonomy with controlled interventions
Humans stay in the loop when systems need help. Interventions should be deliberate, scoped, and auditable.
Minimize per-deployment overhead
Robotics teams shouldn’t rebuild the same data plumbing at every company.
Join us
Careers
We’re looking for exceptional engineers who want to push the boundaries of robotics and physical AI. You’ll work closely with us, own real systems, and ship fast.
Senior Software Engineer, Distributed Systems (Rust)
Full-time · Delft | Zurich | Remote
Build the Rust backend for reliable edge-to-cloud ingestion, indexing, and synchronized retrieval for robot fleets.
Learn more (PDF)Robotics + ML Research Engineer, Data & Models
Full-time · Delft | Zurich | Remote
Run the loop: collect multi-modal robot data, build evaluations, and train models that stress-test and augment DataCore.
Learn more (PDF)We’re building a dynamic, high-trust culture with high ownership and no corporate layers. If you’ve built real systems and care about reliability, reach out.