REFERENCE IMPLEMENTATIONS · DECISIONAL LOGGING

Five regulated segments /
exact integration patterns.

Each example shows a complete POST /api/agents/{uuid}/log call for a real decision type, with the regulatory citations the log entry satisfies. All citations are grounded in primary sources. Copy, adapt, ship.

PRIVACY MODEL — you hash your readable decision locally and send only the content_hash plus non-identifying fields. Readable rationale, inputs, outputs, and evidence never reach Yolo (and are rejected if sent). The examples below use that wire format.
API key requests: agents@yolo.solutions · Endpoint reference: /developers/decisional-logging →
1INSURANCE UNDERWRITING AND CLAIMS

Underwriting AI subject to NAIC, Colorado SB 21-169, and EU AI Act Article 12.

State insurance commissioners are conducting AI examinations now — not as a future liability. The NAIC Model Bulletin on AI Use in Insurance is active in 26+ US states. Colorado SB 21-169's December 2024 deadline has passed; carriers are producing AI governance documentation on demand. EU AI Act Annex III point 5(b) explicitly covers life and health insurance risk assessment and pricing.

COVERAGE DENIAL — CONSEQUENTIAL TIER · $0.01/EVENT
// Coverage denial — event_type: "consequential" ($0.01/event) // // Regulatory obligations satisfied: // NAIC Model Bulletin on AI Use in Insurance — active in 26+ US states (Q1 2026). // Colorado SB 21-169 — December 1, 2024 compliance deadline passed; // carriers are demonstrating compliance to state commissioners now. // EU AI Act Annex III point 5(b) — life and health insurance risk assessment and pricing. // Article 12 record-keeping penalty: 3% of global annual turnover or €15M (Art. 99(4)). // Enforcement: August 2, 2026 (legally operative); December 2, 2027 proposed // via EU Digital Omnibus (May 7, 2026, pending formal enactment). Plan for Aug 2026. // GDPR Article 22 — right to explanation for automated decisions with legal effects. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, // content_hash = SHA-256 you compute over your readable decision bundle // (rationale, inputs, outputs). Those stay on your side; Yolo never sees them. body: JSON.stringify({ event_type: "consequential", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "underwriting-model-v3", model_version: "2026-04-15-prod", correlation_ref: "EXAMPLE-0001", // opaque; SHA-256'd server-side }), } ); const { audit_id, chain_hash, content_hash, verification_url } = await res.json(); // audit_id: 8423 // verification_url: "https://yolo.solutions/verify/8423" // content_hash: "93ca478d...<the hash you submitted, as anchored>...c56a2b0" // // Produce verification_url to the state insurance commissioner on demand. // No Yolo account required to verify; no readable PII ever sent to Yolo.
LIFE OR DISABILITY DECISION — HIGH-STAKES TIER · $0.10/EVENT
// Life or disability decision — event_type: "high_stakes" ($0.10/event) // Evidence by hash manifest — the regulator/plaintiff verifies their retained copy // against the recorded SHA-256. The document itself never leaves your systems. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "high_stakes", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "life-underwriting-v2", model_version: "2026-03-01-prod", evidence_manifest: [ // Hashes only — the physician report stays on your side. { type: "json", sha256: "<64-hex SHA-256 of the physician report>" }, ], correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, content_hash, verification_url } = await res.json(); // audit_id: 8424 // verification_url: "https://yolo.solutions/verify/8424" // content_hash: "<the hash you submitted, as anchored>" // Any auditor verifies their copy of the evidence against the manifest SHA-256.

100 consequential decisions/day + 10 high-stakes/day → ~$60/month per underwriting AI agent. Large carrier with multiple AI systems: $50K–$500K/year under enterprise contract. Insurance segment detail →

2CLINICAL AI · HIPAA + FDA SAMD

Clinical AI under simultaneous HIPAA audit controls and FDA SaMD documentation requirements.

HIPAA 45 CFR §164.312(b) requires audit controls for all systems containing ePHI — a required standard, not addressable. FDA 21 CFR Part 11 §11.10(e) audit trail requirement is the most frequently cited finding in FDA 483 inspections for AI-assisted SaMD. The FDA PCCP guidance (December 2024) mandates documented records for all AI/ML model changes. EU MDR SaMD compliance deadline: August 2, 2028.

DIAGNOSTIC RECOMMENDATION — HIGH-STAKES TIER · $0.10/EVENT
// Diagnostic AI recommendation — event_type: "high_stakes" ($0.10/event) // // Regulatory obligations satisfied: // HIPAA 45 CFR §164.312(b) — audit controls (required standard, not addressable). // 21 CFR Part 11 §11.10(e) — most frequently cited FDA 483 finding for AI-assisted SaMD. // FDA PCCP guidance (December 2024): all AI/ML model changes require auditable records. // EU AI Act Annex III point 5(a) — emergency dispatch triage AI (Article 12, Aug 2026*). // EU MDR 2017/745 — SaMD Article 6(1) path; compliance deadline August 2, 2028 // (Digital Omnibus provisional, pending formal enactment). // // HIPAA-safe by construction: no PHI is sent to Yolo. You hash the readable decision // and each evidence file locally; only the hashes are logged. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "high_stakes", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "ecg-stemi-detector-v4", model_version: "2026-02-15-prod", evidence_manifest: [ { type: "json", sha256: "<64-hex SHA-256 of the model card / scores>" }, ], correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, content_hash, verification_url } = await res.json(); // audit_id: 8425 // verification_url: "https://yolo.solutions/verify/8425" // content_hash: "<the hash you submitted, as anchored>" // FDA reviewers and HIPAA auditors verify via verification_url — no Yolo account, // and no ePHI ever transmitted to Yolo.

50 AI agents per health system at mixed profile → ~$54K/year. The evidence hash manifest satisfies FDA post-market surveillance documentation for AI-assisted SaMD — and the readable model card and ePHI never leave your systems. Healthcare segment detail →

3ALGORITHMIC TRADING · MIFID II RTS 6

Trading AI subject to the most prescriptive existing mandate: MiFID II RTS 6.

MiFID II RTS 6 requires per-algorithm output records, 5-year retention under MiFIR Article 25, and microsecond timestamps for HFT — and firms must be able to reconstruct each algorithmic trading decision from the log. The ESMA Supervisory Briefing on Algorithmic Trading (February 2026, ESMA74-1505669079-10311) confirms active supervisory focus. DORA enforcement began January 17, 2025. Yolo's audit chain has no TTL — nightly Base anchor provides permanent retention.

TRADE DECISION — CONSEQUENTIAL TIER · $0.01/EVENT
// Algorithmic trade decision — event_type: "consequential" ($0.01/event) // // Regulatory obligations satisfied: // MiFID II RTS 6 — per-algorithm output records, 5-year retention (MiFIR Article 25), // microsecond timestamps for HFT. Firms must reconstruct each trading decision from log. // ESMA Supervisory Briefing February 2026 (ESMA74-1505669079-10311) — active focus. // FINRA Rule 3110 — June 2024 regulatory notice confirming AI oversight documentation. // SR 11-7 model risk management — Fed confirmed AI/ML models in scope 2024. // DORA — enforcement January 17, 2025; AI systems in ICT risk management framework. // // Retention: Yolo audit chain anchored to Base nightly — no TTL, satisfies 5-year requirement. // Reconstruction: your retained order/decision bundle hashes to the anchored content_hash. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "consequential", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "momentum-strategy-v5", model_version: "algo-2026-04-01", decided_at: "2026-05-23T13:30:00.123456Z", // microsecond precision (RTS 6) correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, chain_hash, content_hash, verification_url } = await res.json(); // audit_id: 8426 // verification_url: "https://yolo.solutions/verify/8426" // content_hash: "92cb7e77...<the hash you submitted>...d4c07499" // Reconstruction path: content_hash → your retained trade blotter/decision bundle. // Satisfies RTS 6 "ability to reconstruct each algorithmic trading decision."

High-frequency shops generate thousands of consequential events per hour. 10K events/day → $1,000/month. Enterprise contracts available above 1M events/month. Financial AI segment detail →

4CONTENT MODERATION · DSA ARTICLE 34/37

VLOP content moderation under DSA Article 34 systemic risk assessment and Article 37 audit requirements.

DSA Article 34 requires VLOPs (45M+ EU monthly users) to conduct annual systemic risk assessments with 3-year documentation retention — active enforcement. DSA Article 37 mandates annual independent external audits; evidence must be accessible to the European Commission and national DSCs. DSA non-compliance: up to 6% of global annual turnover. UK Online Safety Act: Ofcom launched 5 enforcement programs with 21 active investigations as of May 2026.

SPAM SIGNAL — ROUTINE TIER · $0.0001/EVENT
// High-volume spam signal — event_type: "routine" ($0.0001/event) // // Regulatory obligations satisfied: // EU Digital Services Act (DSA) Article 34 — annual systemic risk assessment documentation // for VLOPs (45M+ EU monthly users). 3-year documentation retention required. // DSA penalty for VLOPs: up to 6% of global annual turnover. // UK Online Safety Act — Ofcom enforcement began 2025; Ofcom launched 5 enforcement // programs, 21 investigations as of May 2026. // // At 100M routine decisions/day: $10,000/day — contact agents@yolo.solutions for bulk rates. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "routine", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "spam-classifier-v8", model_version: "2026-05-01-prod", correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, content_hash, verification_url } = await res.json(); // audit_id: 8427 // verification_url: "https://yolo.solutions/verify/8427"
ENFORCEMENT ACTION — CONSEQUENTIAL TIER · $0.01/EVENT
// Enforcement action — event_type: "consequential" ($0.01/event) // // DSA Article 37 — VLOPs subject to annual independent external audits. // Audit evidence: chain of every enforcement action, 3-year retention, // accessible to European Commission and national Digital Services Coordinators. // DSA Article 17 — users must be notified of enforcement decisions with reasons. // verification_url in response satisfies the "accessible explanation" requirement. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "consequential", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "cib-detector-v3", model_version: "2026-04-20-prod", correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, content_hash, verification_url } = await res.json(); // audit_id: 8428 // verification_url: "https://yolo.solutions/verify/8428" // Present verification_url to the user at appeal — independent of the platform's own logs. // DSA auditor requests the chain for this account cluster: hash-linked, 3-year retention.

100M routine + 1M consequential decisions/day → $20,000/day ($7.3M/year). DSA audit evidence is the purchasing trigger — not a future obligation. Content moderation segment detail →

5GOVERNMENT ELIGIBILITY AI · EU AI ACT ANNEX III

Public administration AI explicitly classified as Annex III high-risk — Article 12 applies.

EU AI Act Annex III point 5 explicitly covers AI systems making benefits eligibility determinations, creditworthiness decisions, and emergency dispatch triage. Article 12 record-keeping penalty: 3% of global annual turnover or €15M under Article 99(4). GDPR Article 22 grants EU data subjects the right to human review of solely automated decisions with legal effects — the verification_url in Yolo's response satisfies the "meaningful information" requirement at contest. Enforcement: August 2, 2026 (legally operative); public authority deployers provisionally extended to August 2, 2030 under Digital Omnibus (pending enactment). AI system providers supplying government deployments face the 2026 date.

BENEFITS DENIAL — HIGH-STAKES TIER · $0.10/EVENT
// Benefits denial — event_type: "high_stakes" ($0.10/event) // // Regulatory obligations satisfied: // EU AI Act Annex III point 5 — access to essential services (benefits eligibility // AI is Annex III high-risk by explicit classification). // Article 12 record-keeping: 3% of global annual turnover or €15M (Art. 99(4)). // General enforcement: August 2, 2026 (legally operative). // Public authority deployers: August 2, 2030 (EU Digital Omnibus provisional, // May 7, 2026, pending formal enactment). Plan for 2026 for AI providers. // GDPR Article 22 — right to human review of solely automated decisions with legal effects. // verification_url satisfies the "meaningful information" requirement on contest. // // Evidence by hash manifest — the appeals tribunal verifies the agency's retained // domain scores against the recorded SHA-256, without trusting the agency's records. const res = await fetch( `https://yolo.solutions/api/agents/${AGENT_UUID}/log`, { method: "POST", headers: { "Authorization": `Bearer ${YOLO_API_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ event_type: "high_stakes", content_hash: "<64-hex SHA-256 of your readable decision bundle>", model_id: "disability-assessment-v3", model_version: "2026-04-01-prod", evidence_manifest: [ { type: "json", sha256: "<64-hex SHA-256 of the domain scores>" }, ], correlation_ref: "EXAMPLE-0001", }), } ); const { audit_id, content_hash, verification_url } = await res.json(); // audit_id: 8429 // verification_url: "https://yolo.solutions/verify/8429" // content_hash: "<the hash you submitted, as anchored>" // Give verification_url to the applicant at time of denial — satisfies GDPR Article 22.

$1M–$50M/year per agency under enterprise contracts. Procurement cycles are 18–36 months — position now for inbound demand when August 2026 enforcement date creates urgency. Government segment detail →

Every entry is independently verifiable.

The verification_url in every response is a public page that shows the full chain entry — content hash, chain hash, preceding entry, and the nightly Merkle anchor transaction on Base. No Yolo account required to view. Any regulator, auditor, plaintiff, or counterparty can verify independently.

Example: https://yolo.solutions/verify/8423 — shows the coverage denial entry from Reference 1 above with its full chain linkage. High-stakes entries carry an evidence hash manifest — anyone can verify their retained copy of the evidence against the recorded SHA-256, with no readable PII held by Yolo.

COMPLIANCE GROUNDING

All regulatory citations in these examples are verified against primary sources. EU AI Act: Annex III categories from the official service desk (ai-act-service-desk.ec.europa.eu). Enforcement dates: August 2, 2026 (legally operative); December 2, 2027 proposed for standalone Annex III systems via EU Digital Omnibus (May 7, 2026), pending formal enactment. Penalties: 3% of global annual turnover or €15M under Article 99(4) for Article 12 violations.

ALL 18 SEGMENT PAGES →FULL API REFERENCE →

API keys, volume pricing, and enterprise onboarding.

agents@yolo.solutions