Evidence over opinion
Every claim we make about a robot is backed by traceable, reproducible data.
About Certisto
We make robot readiness measurable.
Certisto is the certification layer for Physical AI — proving robot policies against faithful twins of the real world and issuing evidence everyone can trust.
Our mission
Every robot that enters the physical world should first be proven in a twin of that world — with evidence anyone can trust.
Our story
We started Certisto because "it worked in sim" kept shipping recalls.
Our founders spent years training robot policies that looked flawless in simulation and then failed in the field. The gap was never a mystery — inaccurate twins, wrong sensor noise, untested edge cases — but no one could measure it, let alone certify against it.
So we built the layer that was missing: a way to prove a policy against a faithful twin of its real environment and issue an artifact everyone — engineers, executives, customers, regulators — could trust.
Today Certisto is building the certification standard for Physical AI, backed by a data flywheel that gets more accurate with every robot deployed.
By the numbers
- Founded
- 2026, San Francisco
- Category
- Physical AI · Robotics validation
- Focus
- Sim-to-real certification
- Backing
- Deep-tech & robotics investors
What we believe
Principles that keep a certificate worth trusting.
Calibrated honesty
We report confidence bounds and failure modes, not just the numbers that look good.
Safety is the product
A certificate that isn't trustworthy is worse than no certificate at all.
Compounding truth
Every deployment makes the next prediction sharper. The flywheel is the moat.
Partner, don't gatekeep
We win when our customers ship confidently — not when we slow them down.
Standards, openly built
The certification benchmark only matters if the whole industry can trust it.
2026
Year founded
3
Product lines shipping
9
NVIDIA technologies integrated
100%
Certificates fully auditable
Milestones
How we're building the certification layer.
Q1 2026
Company founded
Certisto forms around a single idea: turn simulation into an auditable readiness certificate.
Q2 2026
Design-partner program
First robotics teams validate humanoid and industrial-arm policies on the platform.
Q3 2026
Readiness benchmark published
We release the sim-to-real correlation study to seed an open certification standard.
Q4 2026
Edge runtime GA
Jetson conformance monitoring closes the loop from certificate to field telemetry.
Leadership
A team from robotics, simulation, and safety.
HR
Hussain Raza
Founder & CEO
Former robotics autonomy lead. Shipped manipulation policies to production fleets.
JM
Jonas Meyer
Co-founder & CTO
Simulation and RL researcher. Built large-scale parallel training infrastructure.
PV
Priya Varma
Head of Safety
Functional-safety engineer for autonomous systems and industrial robotics.
TK
Tomas Kite
Head of Product
Enterprise deep-tech product leader focused on developer and evidence workflows.
Backed by
Investors who understand deep tech and robotics.
Horizon DeepFoundry CapitalKinetic VenturesNorth Robotics FundEmbodied Partners
Working here
Rigor, ownership, and a bias for evidence.
- Remote-first with hubs in San Francisco and Berlin
- Deep-work culture — we protect maker time
- Competitive equity and comprehensive benefits
- Real hardware, real robots, real deployment problems
Impact you can measure
Your work directly decides whether a robot is safe to ship.
Frontier problems
World models, sim-to-real estimation, and GPU-scale validation.
Small, senior team
Low process, high trust, and direct ownership of outcomes.
"The most interesting thing in Physical AI isn't a new robot — it's a way to prove the robots we already have are ready."
Robotics Weekly
Help us certify the physical world.
Whether you want to validate your robots or build the platform, we'd love to talk.