Identity resolution + write guard

Resolve identity before every write

Anchord is the read-only identity + trust layer agents use before writing to HubSpot, Salesforce, Stripe, or any other system. Resolve records to canonical identities, guard every write, and escalate ambiguity to humans — never guess, never duplicate.

Your first resolve call

Sign up, get your API key, and run these two commands. That's it.

# 1. Verify your key
curl -s https://api.anchord.ai/api/ping \
  -H "Authorization: Bearer YOUR_KEY"

# 2. Resolve a company
curl -s -X POST https://api.anchord.ai/api/v1/resolve/company \
  -H "Authorization: Bearer YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain": "acme.com", "name": "Acme Corp"}'

Replace YOUR_KEY with the API key from signup. The resolve call returns resolved, not_found, or needs_review.

I'm building an agent

Set up the MCP server for Cursor or Claude Desktop, or call the REST API directly. Resolve identity and guard writes in your agent's workflow.

Agent Quickstart →

I'm setting up Anchord for my team

Connect your CRM, explore your data, and manage identity quality from the console. Load demo data to see the product immediately.

Full Quickstart →

Why agents need an identity layer

Agents can't verify identity

An agent sees a name and domain. It can't tell whether that's an existing record, a duplicate, or a new entity without a canonical reference.

CRMs don't prevent cross-system dupes

Salesforce and HubSpot catch duplicates within their own system. Neither sees records in the other, or in Stripe, or in your data warehouse.

Guessing creates data debt

Every wrong write — a duplicate contact, an update to the wrong account — compounds into data quality problems that are expensive to fix later.

How it works

Connect source records, resolve to an AnchorID, evaluate proposed writes — three API steps before your agent touches the external system.

1

Ingest

Push records from any source via the batch API, or connect HubSpot, Salesforce, and Stripe through built-in OAuth integrations.

2

Resolve

Match the record to an AnchorID — Anchord's canonical identity record for a Company or Person — using rules-based scoring. Get back resolved, not_found, or needs_review.

3

Guard

Before writing, evaluate the write against Anchord. Get allowed or blocked with specific block codes. Anchord never writes — the caller decides.

When identity is ambiguous, agents stop

Most identity systems guess when there are multiple plausible matches. Anchord returns needs_review instead.

Agent gets needs_review

Two or more candidates scored too close to auto-resolve. The agent does not write.

Human reviews in console

The Review Queue shows incoming data, candidates with scores, and a recommendation. One click to confirm, dismiss, or create new.

Agent retries cleanly

After human resolution, the same resolve call returns resolved with a stable AnchorID.

When to use Anchord

Any time your agent or pipeline is about to write to an external system. See all scenarios →

Read-only (no writebacks)

Anchord never writes to your systems. Guard returns a decision — the caller executes.

Encrypted at rest & in transit

AES-256, TLS 1.2+, secrets in AWS Secrets Manager.

Tenant-isolated

Every query scoped to your tenant. No cross-tenant access paths.

Audited operations

Merges, key events, and sensitive actions logged with request IDs.

Security details · Privacy Policy · Contact us

Start building

Sign up, get your API key, and run your first resolve call. Free plan, no credit card.