Built for teams where AI decisions have real consequences.
Governance, auditability, and compliance infrastructure for organizations operating AI in regulated or high-stakes environments.
EU AI Act · Colorado AI Act · GDPR Compatible · Air-gap Ready · SOC 2 Aligned
Tamper-evident audit trails
Every agent run produces a lineage record signed with HMAC-SHA256 using your project secret key. Records are tamper-evident, queryable, and exportable for regulators.
Run ryva lineage verify --all to verify that no lineage record has been altered since it was written. Any tampering is detected immediately.
Records are indexed by run ID, agent name, timestamp, and parent run ID. You can reconstruct the complete decision history for any agent output, including multi-step pipelines.
$ ryva lineage verify --all
Verifying 142 lineage records...
Records verified: 142 / 142
Tampered: 0
Signature algo: HMAC-SHA256
Key ID: corp-key-2026-01
All 142 records verified
No tampering detected
EU AI Act and Colorado AI Act compliance
Continuous documentation, not annual. Every command updates your compliance evidence in real time.
| Article | Requirement | Ryva Coverage | Status |
|---|---|---|---|
| Article 9 | Risk management system | Structured test results, lineage history, governance scores | ✓ Covered |
| Article 10 | Data governance | PII masking, retrieval source recording, data provenance | ✓ Covered |
| Article 12 | Record-keeping | Tamper-evident lineage records, HMAC signatures, exportable | ✓ Covered |
| Article 13 | Transparency | Auto-generated model cards, alignment rule declarations | ✓ Covered |
| Article 14 | Human oversight | Alignment failure blocks, high-risk run flagging | ✓ Covered |
| Article 15 | Accuracy and robustness | Fuzz, adversarial, and regression testing on every push | ✓ Covered |
Complete audit package in one command.
One command generates a complete audit zip containing your governance report, model cards, verified lineage records, and compliance checklists. Hand it directly to your legal team.
The package includes both EU AI Act and Colorado AI Act checklists, all 142 verified lineage records, and a machine-readable manifest regulators can process programmatically.
$ ryva audit export
✓ Audit package ready
ryva_audit_myproject_20260529.zip
├── README.md
├── governance/
│ ├── governance_report.json
│ └── governance_report.md
├── model_cards/
│ └── intake_agent_model_card.json
├── lineage/
│ └── (142 records, all verified)
├── compliance/
│ ├── eu_ai_act_checklist.md
│ └── colorado_ai_act_checklist.md
└── PACKAGE_MANIFEST.json
Your first compliance audit. 30 days.
We connect to your existing AI stack, assess against EU AI Act and Colorado AI Act requirements, and deliver a complete audit package.
What you get
Stack assessment
We inventory all AI systems and map them to regulatory requirements.
Evidence generation
We run Ryva against your systems and generate all compliance documentation.
Audit package
Complete zip file with everything your legal team needs for regulators.
Every organization is different.
Book a discovery call and we will scope exactly what your team needs.
Book a discovery callGet in touch
We respond within one business day.