Pre-Seed 2026
The Execution &
Evidence Layer for AI
We run your AI safety controls and prove they executed. Third-party witnessed, zero data egress.
The Problem
AI is running in production. Unwitnessed.
Every AI interaction in regulated industries carries risk—right now. When regulators, auditors, or plaintiffs come asking, can you prove what happened?
How do we know your guardrails actually ran on every inference?
Can you prove consent was captured before each AI interaction?
What happens when the plaintiff's attorney asks for evidence?
Your dashboard says "compliant."
Their lawyers say "prove it."
Nov 2025: Sharp HealthCare class action alleges ambient scribe hallucinated consent for ~100,000 patients. No cryptographic proof of what controls ran.
What's Happening Now
Every industry. Every interaction. Unwitnessed risk.
Sharp HealthCare Class Action
~100k patients. Ambient scribe faked consent.
Sales AI Under Scrutiny
Training AI with pre-release drug data requires provable boundaries.
OCC Model Risk Guidance
SR 11-7 requires audit trails for AI-assisted decisions.
Claims AI Challenged
Regulators demand explainability proof for denial decisions.
The pattern: AI is running in production right now. When something goes wrong, there's no witness, no proof of what controls ran. Just logs that lawyers tear apart.
What We Do
Witnessed AI deployment
Every AI interaction gets a cryptographic witness—timestamped proof that safety controls executed. When regulators, auditors, or plaintiffs ask what happened, you have evidence they can verify.
When the attorney asks "prove it"—you can. When the procurement committee asks for evidence—you have it. When the auditor arrives—you're ready. And after the deal closes, you have ongoing proof for internal compliance, board reporting, and liability protection.
Why Now
The liability clock is ticking
Every day without witnessed deployment is another day of accumulating risk. Regulations are coming—but lawsuits and audits aren't waiting.
COMPETITIVE ADVANTAGE
Close faster: Answer security questionnaires definitively
Reduce friction: Give governance committees what they need
Differentiate: Your competitors can't prove execution. You can.
REGULATORY DEADLINES
Standard of care precedes law. NIST AI RMF and ISO 42001 are already the de facto duty of care.
The window: In 18 months, AI vendors without cryptographic proof will lose deals to those who have it.
The Insight
No one else proves execution
Guardrails block bad things. GRC platforms document policies. Neither proves your controls actually ran on every inference.
Infrastructure Players
Cloudflare, AWS, Alinia
Block bad things at runtime
Gap: No compliance mapping. Logs ≠ proof.
GRC Platforms
Vanta, Credo AI
Document policies and procedures
Gap: No runtime enforcement. Trust us.
GLACIS
Cryptographic Attestation
Prove execution at inference time
Runtime + Compliance + Proof
Hyperscalers can't witness their own AI. We're the independent third party.
The Platform
Create attestations. Monitor compliance.
A single pane of glass for AI governance—track every attestation, map controls to frameworks, and generate audit-ready evidence.
Traction
Strong pull. Just getting started.
Next milestone
JPM Healthcare 2026
40k+
visits/day
Colorado-based. Facing Colorado AI Act deadline.
Consent attestation • PHI proof • Guardrail evidence
Tenant-bounded AI with trade secrets, pre-release drug data.
Query attestation • Tenant isolation • Trade secret fencing
Also in pipeline
Inbound interest
Health Systems
AI governance inquiries
AI Vendors
Procurement pain resonates
Investors
Regulated AI + governance thesis
Validation: MultiCare AI Governance Committee blocked Ambience deal over data ownership / chain-of-custody concerns.
Team
FDA Authorized. Enterprise Deployed.
We've lived this problem.
Joe Braidwood
Co-Founder & CEO
SwiftKey → 1 in 4 smartphone users
Founding exec, $250M Microsoft exit. Chief Strategy at Vektor Medical—secured reimbursement for AI device. Cambridge Law.
Dr. Jennifer Shannon
Co-Founder & CMO
Cognoa → First FDA De Novo for AI diagnostics
Medical Director at Cognoa. Navigated FDA authorization for AI that diagnoses autism in children.
Rohit Tatachar
CTO
Microsoft Azure → $2B product line
Engineer turned product leader. Led Azure's enterprise platform business. Personal relationships with top 50 Azure customer CEOs.
Caer
Rust & Cryptography Lead
10+ years production • WPI
Atreya
AI Engineer
Model eval & red teaming • WPI • Ex-PAAI
Advisors
4 Families
Patent portfolio filed
Fenwick & West • Nov 2025
Defensibility
Building the Category Definition
Co-Epoch Attestation
Binds receipts to specific binary versions. Prevents "binary substitution" attacks.
Statistical Safety Signals
Mathematically rigorous sampling for continuous compliance verification.
Zero-Payload Egress
Cryptographic commitments without data exposure. PHI never leaves boundary.
Federated Witness Network
Distributed attestation for independent verification at scale.
Filed Nov 2025 — First-mover advantage in attestation-based AI compliance.