About GLACIS

Runtime assurance for AI
systems that act.

Glacis is the runtime proof layer underneath your AI. We build software that classifies locally, enforces at the action boundary, and produces verifiable receipts — so your team can defend the trust they’ve been asked to stand behind.

  1. See runtime visibility
  2. Control local guards
  3. Prove signed receipts
One workflow, local controls, portable proof.
What We Build

See. Control. Prove. Improve.

AI systems are moving from assistance to action. Most teams can describe policies, logs, and evals; few can produce a portable record that proves which runtime controls actually ran when a workflow made a consequential move.

See — map the runtime surface

We start with the workflow: prompts, tools, data access, human approvals, escalation paths, and failure modes. The result is a concrete map of where an agent can act and where proof is required.

Control — harden locally

Runtime controls sit inside your environment, with classifiers running on your own hardware. Sensitive prompts, PHI, secrets, and customer data never leave your stack to decide whether they should — the soft spot in every API-based safety tool.

Prove — generate signed receipts

Every important control event can produce a signed receipt: what control ran, what decision it made, which workflow it protected, and how a reviewer can verify integrity. Receipts roll up into evidence packs for security, compliance, buyers, and internal governance.

Improve — close the assurance loop

Evidence should not be a dead archive. It should show where controls are catching risk, where reviewers still intervene, and where the next workflow hardening pass will have the most leverage.

50+
Combined years in AI
$250M
Exit (SwiftKey → MSFT)
1B+
Devices shipped to
FDA
Clearance experience
Leadership

We lived the problem

We built the “ship fast vs. prove it works” paradox at SwiftKey, Vektor Medical, and Cognoa. We’re not outsiders guessing — we built the systems that needed this infrastructure.

Joe Braidwood

Joe Braidwood

Co-Founder & CEO

Founding team at SwiftKey (acquired by Microsoft for $250M), scaled the keyboard AI to 300M+ devices pre-acquisition; the technology now ships on 1B+ devices globally. Chief Strategy Officer at Vektor Medical. 15+ years building AI products at scale.

Consumer AI Scale Healthcare GTM Cambridge Law
LinkedIn
Rohit Tatachar

Rohit Tatachar

Co-Founder & CTO

Nearly 19 years at Microsoft leading Azure infrastructure at billion-dollar scale. Most recently on the AI Foundry team building the infrastructure layer for enterprise AI. Architected the Glacis enforcement stack from first principles.

Azure Infrastructure AI Foundry Hyperscale
LinkedIn
Dr. Jennifer Shannon

Dr. Jennifer Shannon, MD

Co-Founder & Chief Medical Officer

Board-certified psychiatrist with 15+ years clinical experience. Medical Director at Cognoa — helped develop Canvas Dx, the first FDA-authorized AI diagnostic for autism. CHAI Coalition member.

FDA Clearance Clinical AI Digital Health
LinkedIn
Principles

What we believe

Private by design

We architect systems so we never see your data. Trust through verification, not promises.

Built for builders

Accountability that slows you down isn’t accountability — it’s a tax. We optimize for both speed and rigor.

Prove, don’t promise

Attestations over assertions. Verifiable evidence over policy documents.

Trust you can defend

We’re underneath the dashboard, not the dashboard itself. The evidence we produce belongs to your team — to hand to auditors, regulators, customers, and reviewers, and have them verify it independently.

If you’re trying to figure out where you stand with AI — security, customer review, regulators, or just your own internal pressure — we’d like to hear about it. No commitment, no script.

Talk to the team
Careers

Join us

We’re building the runtime assurance layer for the AI economy. If that sounds interesting, we’d love to hear from you.