ProductUse CasesDevelopersCompanyBlogRequest Access

From AI decision to compliance evidence. In every regulated context.

Aegis Trace provides complete decision provenance for every regulated industry where AI systems influence outcomes that must be evidenced. The same API, the same certificate format, the same tamper-proof audit trail.

Use Case 01 , Financial Services

A wealth manager's AI recommends selling a position. The FCA asks why.

The situation

A client's portfolio is reviewed by an AI-assisted suitability engine. It recommends reducing exposure to a fund based on risk profile, market conditions, and client objectives. Six months later, the FCA requests evidence that the recommendation was suitable under Consumer Duty. Without Aegis Trace, the firm has application logs. With Aegis Trace, they have a signed certificate with complete decision provenance.

The regulations

FCA PS22/3Consumer DutyMiFID IIEU AI Act Art.12
POST /v1/decisions
{
  "agent_id": "suitability-engine-v3",
  "decision_type": "portfolio_recommendation",
  "input_hash": "sha256:a3f8c1d...",
  "output": {
    "action": "REDUCE_EXPOSURE",
    "fund": "GB00B3X7QG63",
    "rationale": "Risk profile mismatch, client objective: capital preservation"
  },
  "regulatory_context": ["FCA_PS22_3", "CONSUMER_DUTY"],
  "client_ref": "[REDACTED-BY-PRESIDIO]"
}
HTTP 201 CreatedVERIFIED
{
  "certificate_id": "AT-2026-03-22-f8a3c1d",
  "verdict": "PASS",
  "confidence": 0.9847,
  "hmac_signature": "a3f8c1d9e2b4f...",
  "pii_redacted": true,
  "regulatory_refs": ["FCA PS22/3", "Consumer Duty"],
  "sealed_at": "2026-03-22T11:53:52Z",
  "retention_until": "2033-03-22T00:00:00Z"
}
What the FCA receives: A tamper-proof certificate with complete decision provenance, from data input through model inference to final output, retrievable via API in under 200ms.

Use Case 02 , Healthcare & Life Sciences

A clinical decision support system flags a drug interaction. The MHRA needs the audit trail.

POST /v1/decisions
{
  "agent_id": "clinical-dss-v2",
  "decision_type": "drug_interaction_flag",
  "input_hash": "sha256:b7e2a9f...",
  "output": {
    "flag": "INTERACTION_WARNING",
    "severity": "HIGH",
    "drug_pair": "[REDACTED]",
    "recommendation": "REVIEW_BEFORE_PRESCRIBING"
  },
  "regulatory_context": ["MHRA_AI", "EU_MDR"],
  "patient_ref": "[REDACTED-BY-PRESIDIO]"
}
HTTP 201 CreatedVERIFIED
{
  "certificate_id": "AT-2026-03-14-c9d2e8f",
  "verdict": "PASS",
  "oversight_required": true,
  "pii_redacted": true,
  "regulatory_refs": ["MHRA AI Guidance", "EU MDR Art.61"],
  "sealed_at": "2026-03-14T09:22:11Z",
  "retention_until": "2036-03-14T00:00:00Z"
}

The situation

An AI system assists a clinician by flagging a potential drug interaction based on a patient's prescription history. The recommendation influences the prescribing decision. Under MHRA AI regulations and the EU MDR, the hospital must demonstrate the AI's recommendation was within validated parameters and was appropriately overseen. Aegis Trace provides complete decision provenance for every clinical AI output.

The regulations

MHRA AI GuidanceEU MDR Art.61NHS AI FrameworkGDPR Art.22
What the MHRA receives: A complete decision record with validated model version, redacted patient reference, severity classification, and cryptographic seal, structured for regulatory submission.

Use Case 03 , Insurance & Underwriting

An underwriting model declines a policy. The applicant requests an explanation under GDPR Art.22.

The situation

An automated underwriting model assesses a commercial insurance application and returns a decline decision. Under GDPR Article 22 and FCA ICOBS, the applicant has the right to request a meaningful explanation of the automated decision. The insurer must produce a record of what the model assessed, what it decided, and on what basis. Aegis Trace captures the full decision provenance automatically.

The regulations

GDPR Art.22FCA ICOBSSolvency IIEU AI Act Art.12
POST /v1/decisions
{
  "agent_id": "underwriting-model-v4",
  "decision_type": "policy_assessment",
  "input_hash": "sha256:d4c1b8e...",
  "output": {
    "decision": "DECLINE",
    "risk_score": 0.847,
    "primary_factors": ["claims_history", "sector_risk"],
    "explanation_ref": "EXP-2026-03-11-449"
  },
  "regulatory_context": ["GDPR_ART22", "FCA_ICOBS"],
  "applicant_ref": "[REDACTED-BY-PRESIDIO]"
}
HTTP 201 CreatedVERIFIED
{
  "certificate_id": "AT-2026-03-11-d4c1b8e",
  "verdict": "PASS",
  "explanation_available": true,
  "pii_redacted": true,
  "regulatory_refs": ["GDPR Art.22", "FCA ICOBS", "Solvency II"],
  "sealed_at": "2026-03-11T14:07:33Z",
  "retention_until": "2033-03-11T00:00:00Z"
}
What the applicant and regulator receive: A structured explanation record tied to a tamper-proof certificate, satisfying the GDPR Article 22 right to explanation and FCA ICOBS evidencing requirements.

Use Case 04 , AI Credit Decisioning

Your AI credit model declined a mortgage application. The borrower appeals.

POST /v1/decisions
{
  "agent_id": "credit-scoring-v2",
  "decision_type": "mortgage_assessment",
  "input_hash": "sha256:e7f3b2a...",
  "output": {
    "decision": "DECLINE",
    "risk_score": 0.782,
    "primary_factors": [
      "debt_to_income_ratio",
      "employment_tenure"
    ],
    "explanation_ref": "EXP-2026-04-02-891"
  },
  "regulatory_context": ["FCA_CONC", "GDPR_ART22"],
  "applicant_ref": "[REDACTED-BY-PRESIDIO]"
}
HTTP 201 CreatedVERIFIED
{
  "certificate_id": "AT-2026-04-02-e7f3b2a",
  "verdict": "PASS",
  "explanation_available": true,
  "pii_redacted": true,
  "regulatory_refs": ["FCA CONC", "GDPR Art.22"],
  "sealed_at": "2026-04-02T10:18:44Z",
  "retention_until": "2033-04-02T00:00:00Z"
}

The situation

An AI-assisted credit scoring model evaluates a mortgage application and returns a decline. Under FCA CONC and GDPR Article 22, the firm must provide a meaningful explanation of the automated decision and demonstrate the assessment was fair. Without Aegis Trace, the firm has model logs scattered across systems. With Aegis Trace, there is a single, signed decision record with complete provenance.

The regulations

FCA CONCGDPR Art.22EU AI ActSR 11-7
What the borrower and regulator receive: A single, tamper-proof decision record with every input, risk factor, and model output. Every AI decision with full provenance, traceable, verifiable, defensible.

Use Case 05 , Algorithmic Trading

Your trading algorithm executed a series of orders during market volatility. The FCA requests a reconstruction.

The situation

An algorithmic trading system made autonomous decisions during a period of market stress. Under MiFID II and FCA SYSC 9, the firm must be able to reconstruct the decision chain: what data the model consumed, what signals it acted on, and what orders it placed. Aegis Trace provides a sealed, time-stamped record for every execution decision, with complete decision provenance from market input to order output.

The regulations

MiFID IIFCA SYSC 9MAR
POST /v1/decisions
{
  "agent_id": "algo-trading-v7",
  "decision_type": "execution_order",
  "input_hash": "sha256:f9a2c7d...",
  "output": {
    "action": "SELL",
    "instrument": "GBPUSD",
    "quantity": 500000,
    "signal_strength": 0.91,
    "market_condition": "HIGH_VOLATILITY"
  },
  "regulatory_context": ["MIFID_II", "FCA_SYSC_9"],
  "execution_ref": "ORD-2026-04-07-1142"
}
HTTP 201 CreatedVERIFIED
{
  "certificate_id": "AT-2026-04-07-f9a2c7d",
  "verdict": "PASS",
  "pii_redacted": false,
  "regulatory_refs": ["MiFID II", "FCA SYSC 9", "MAR"],
  "sealed_at": "2026-04-07T11:42:08Z",
  "retention_until": "2033-04-07T00:00:00Z"
}
What the FCA receives: A complete, sealed reconstruction of every algorithmic trading decision. Time-stamped, cryptographically signed, and exportable in MiFID II format.

The regulation changes. The requirement does not.

A decision was made
An AI system produced an output that influenced a regulated outcome.
Evidence is required
A regulator, auditor, or court asks: what produced that output, when, and was it appropriate?
Aegis Trace provides it
A tamper-proof certificate with complete decision provenance exists for every decision. Retrievable in under 200ms. Exportable for submission.

Tell us your regulatory context.

We will confirm whether Aegis Trace covers your specific requirements and provide relevant technical documentation.