Build with us
Help us build the models that move the physical world.
A platform running in production and an open research frontier ahead of it. Autonomy where the model has to be right, because the machine moves when it decides.
We already have years of deployed industrial infrastructure, so Reality Models learn continuously from real-world observation rather than only from internet-scale video and simulation.
Reality Model Team · 7 roles
Not a manager who happens to supervise researchers. An active scientist with the credibility to attract other scientists.
Principal Reality Model Research Engineer
Design the architecture of the model itself. Beyond today's transformer-centric approaches: persistent memory, long-horizon forecasting.
Build the machinery that teaches Reality Models. Dataset creation at scale, distributed training, curriculum learning, reproducibility.
Generate universes to learn from. Physics simulation, digital twins, procedural environments, domain randomization.
Answer one question: is the model actually understanding reality? New benchmarks for object permanence, causality, occlusion recovery, physical plausibility.
Reality Infrastructure Engineer
Make trillion-token training possible. GPU clusters, pipeline parallelism, high-throughput dataloaders, cost.
Close the loop between prediction and reality. Connect the model to robots, industrial machines, cameras, PLCs, and autonomous systems.
Spatial Intelligence Team · 3 roles
Computational Geometry Engineer
The geometric foundations that let machines reason about three-dimensional space. Meshes, point clouds, CAD, collision, real-time.
Turn raw sensor data into representations of the physical world. Cameras, LiDAR, radar, IMUs, and industrial sensors, fused and robust under noise.
Make complex physical systems legible to engineers and operators. Real-time digital-twin tooling, model introspection, 2D and 3D interfaces.
Dev Ops Team · 5 roles
VP of Platform Engineering & Information Security
Own the engineering platform behind Reality Models, Vitesse products, cloud, and edge — from research clusters to industrial deployments.
Platform Infrastructure (IT) Engineer
Design and operate the infrastructure researchers and production systems build on. Kubernetes, GPU compute, infrastructure as code, observability.
Quality Engineering & Release Automation Engineer
Automated systems that measure and validate quality continuously — including PLC, robot, and hardware-in-the-loop testing.
Protect the IP, the research infrastructure, customer data, and deployed systems. Product and infrastructure security, detection, and compliance.
The runtime that deploys and operates Movendi software across thousands of edge devices — disconnected, resource-constrained, and mission-critical.
Vitesse Product Development · 3 roles
Interfaces that let engineers and operators visualize, configure, and control physical systems. Dashboards, HMIs, statechart editing, diagnostics.
The services behind orchestration, digital-twin synchronization, telemetry pipelines, and high-volume industrial messaging.
Industrial Connectivity Engineer
The layer that talks to industrial equipment. For someone unafraid to hack into — and break — a machine to understand it better than the people who built it.
Deployment Team · 2 roles
Deployment frameworks, reusable integration components, and the customer-specific work that becomes platform capability.
Make every deployment deliver measurable value, and build the in-house test hardware that mirrors what customers run. Expect to travel.
How to apply
Open the role and apply from there, or write to us directly. Name the role and show us something you have built, shipped, or written. The platform is in production, so your work lands on real equipment — and you find out quickly when it is wrong.