TrustVendor
Book a demo
product

Vendor Pulse API

Ship agent-grade vendor intelligence in your own product — the same claims and citations your workspace uses, available by API.

Book a demo See pricing

What it is

Vendor Pulse API — in detail.

The Vendor Pulse API gives GRC platforms, compliance tools, and procurement applications access to the same agent-produced claims and citations the Vendor Risk workspace runs on. One resolve call turns a domain name or company name into a canonical vendor identity. One pulse call returns posture score, assurance score, active certifications, open signals by severity, and evidence links — every field traceable to an agent-produced claim with an artifact ID and a sha256 hash.

Webhooks fire on agent-produced events: vendor.ready when the Extractor agent finishes profiling a new vendor, score.changed when a score computation runs after new claims land, signal.raised when the Triage agent marks a change as material. Your integration receives the full event payload including claim_ids and evidence_ids — so your customers can click through to the same span-level evidence the workspace shows.

Resolve is permanently free and unmetered. Pulse calls are priced per resolved vendor per month so your integration costs scale with actual usage. Enterprise partners can white-label the entire intelligence layer — no TrustVendor branding, custom domain for snapshot links, and the full agent pipeline operating as invisible infrastructure behind your product.

How it works

Three steps,
fully auditable.

Resolve 01

POST a domain or company name to /v1/vendors/resolve. The Resolver agent returns a canonical vendor_id with confidence score in under 100ms for known vendors. Batch resolution for onboarding an existing book of business accepts up to 1,000 domains in a single call.

Pulse 02

GET /v1/vendors/{vendor_id}/pulse returns posture score, assurance score, certifications with expiry dates, open signals by severity, and evidence links. Every score and certification links to the agent-produced claim that backs it. New vendors return 202 with a status poll endpoint while the Extractor agent runs — typically 2 to 8 minutes.

Subscribe 03

Register a webhook for agent-produced events: vendor.ready, score.changed, signal.raised, certification.expiring, subprocessor.changed. Your integration receives real-time pushes with claim_ids and evidence_ids attached — no polling, no stale data.

What you get

Built for teams that
have to prove things.

Free resolve, metered pulse

Identity resolution is never charged or rate-limited. Pulse calls are priced per resolved vendor per month — $0.40 to $1.20 depending on volume — so your integration costs scale with actual usage.

Webhook-first, agent-event-driven

Webhooks fire on agent-produced events with claim_ids and evidence_ids in the payload. Your customers see changes when agents detect them, not on your next polling cycle.

White-label option

Enterprise API partners can white-label the full agent intelligence layer. Your customers see your brand; TrustVendor operates as infrastructure. Available with a platform partnership agreement.

What the agents will never do

Agent guarantees,
in plain terms.

Agents produce claims, never numbers. Scoring is deterministic Go arithmetic over those claims — models never run in the scoring path.

Every claim cites a hashed, immutable snapshot span. No claim is written to the database without an artifact_id and a validated quote hash.

Extractor agents are prompt-versioned. Every claim records the prompt file, model, and schema version that produced it. A regression is a rollback, not a data loss.

The Analyst agent's system prompt forbids it from asserting a fact it cannot cite. Warnings surface in the response envelope when tool results are incomplete — the answer is never silently filtered.

Sample

What it looks
like in practice.

Request

# 1. Resolve a vendor identity
curl -X POST https://api.trustvendor.co/v1/vendors/resolve \
  -H "Authorization: Bearer tv_live_..." \
  -H "Content-Type: application/json" \
  -d '{"domain":"acme.com"}'

# 2. Fetch the live pulse
curl https://api.trustvendor.co/v1/vendors/vnd_8kQ2mR/pulse \
  -H "Authorization: Bearer tv_live_..."

Response

{
  "vendor_id": "vnd_8kQ2mR",
  "posture": {
    "score": 84,
    "band": "strong"
  },
  "assurance": {
    "score": 71,
    "band": "established"
  },
  "certifications": [
    {
      "framework": "SOC2_TYPE2",
      "status": "active",
      "expires": "2026-11-30",
      "evidence_url": "https://..."
    }
  ],
  "open_signals": {"high": 0}
}

Common questions

Answered plainly.

How do I get started without committing to a volume contract?
The resolve endpoint is permanently free with no account required. Pulse access starts with a sandbox key that covers 100 pulse calls per month — enough to build and test your integration. Production keys and volume pricing are activated when you are ready to launch.
What does "async by default" mean for new vendors?
When you resolve a vendor that TrustVendor has not yet profiled, the resolve call returns 202 with a status of "enriching" and a job ID. You can poll the status endpoint or register a vendor.ready webhook to be notified when the Extractor agent has finished. Enrichment typically takes 2 to 8 minutes for a vendor with a public trust centre.
Is the API versioned?
Yes. The URL carries the major version (/v1). Minor, backwards-compatible changes are announced via the TV-Version header. You can pin to a specific minor version to prevent unexpected response shape changes during your development cycle.
What is included in the evidence links in the API response?
Each certification and signal in the pulse response includes an evidence_url field that links to the specific artifact and claim in the TrustVendor evidence archive. The link is publicly accessible for public-graph data and requires workspace authentication for tenant-uploaded documents.
Which model tier runs which extraction?
Tier-0 is deterministic Go parsers handling structured feeds, JSON schemas, and probes — about 65% of volume at zero inference cost. Tier-1 small models handle relevance filtering, deduplication, and classification — about 25% of volume. Tier-2 mid models handle trust-centre extraction, diff interpretation, and materiality — about 9%. Tier-3 frontier models handle SOC 2 exception analysis, entity adjudication, and portfolio Q&A — about 1%. The routing decision is a pure Go function made before any model call.

See it live

Vendor Pulse API
on your vendors.

Book a 30-minute demo. We’ll run it live on vendors from your register — not demo data.

We will respond within one business day.