Now: roar + GLaaS
Implicit observation & structural reproducibility
- Track what actually ran
- Lookup artifact lineage by its hash
- Reproduce models without archaeology
AI Process Power—
Starting with Lineage
TReqs is the platform for developing AI better together.
We’re starting by making lineage automatic—with roar + GLaaS .
How TReqs Comes to Life
Next: TReqs Platform
Asynchronous collaboration & resource control
- Training Requests, like pull requests
- Shared visibility into compute, cost, and results
- Decisions anchored in real runs
Later: Organization-scale AI Process Power
Making it easy to do the right thing
- Audits become trivial
- The happy path is the compliant path
- Teams move fast together
TReqs Leads AI Best Practices with TRAC
TRAC: Four essential principles of AI Process Power
Traceable
Decisions, changes and data leave trails that teams can follow.
Reproducible
Ready recipes for your special sauce so you can make it again.
Attributable
Inputs, work, and spend are visible.
Collaborative
Shared context replaces meetings and guesswork.
TReqs Helps You Flex Your AI Process Power
Faster cycles
Shorten time‑to‑signal with clearer gates and fewer back‑and‑forths.
Lower spend
Cut wasted runs and chase the right experiments. Your budget, your rules.
Better models
Cleaner inputs and consistent process yield more reliable performance.
Team
Chris Geyer
Co‑founder & CEO
25+ years Robotics, AI & Computer Vision leader; ex-iRobot and previously first engineer at Berkshire Grey building enterprise robotics systems.
Jon Geyer
Co‑founder & CEO
Leader of UX and product design for complex systems; shipping crisp, human‑centered workflows for e-commerce & government science.
Trevor Basinger
Founding Engineer
Deep experience in data infrastructure and applied ML; AI systems builder at Manifold and led the Machine Intelligence division at Flogistix.
Prof. Christian Kästner
Advisor
CMU professor specializing in software engineering for ML systems, open-source sustainability, and supply-chain security.
Rachel Lomasky
Advisor
Product and AI/ML leadership across B2B; pragmatic discipline on outcomes and delivery.
Pilots Open Now
Launching Q1 2026. Limited early‑access slots for teams ready to quantify outcomes.