Product

Reconciliation, built API-first.

Send payment data as part of your existing integration — we normalize, structure, and reconcile it. Pull differences, AI-proposed resolutions, tickets, and comparisons back via API, and build whatever your team needs on top, kept entirely inside your own stack.

Ingest API

POST records from any source — ledgers, schemes, banks — no file formats to agree on

Normalize

Every payload is structured against a common schema the moment it lands

Reconcile Engine

Matching runs continuously in the background — no polling, no batch jobs to schedule

Differences API

Query unresolved breaks, filtered, paginated, and ready to render in your own UI

AI Resolutions

Suggested causes and fixes come back as structured fields, not free text to parse

Webhooks & Events

Subscribe to match, break, and resolution events — build on top in real time

Send

One call in, wherever payments already touch your stack.

Wire the ingest API into the same place you implement payments — a webhook handler, a settlement job, a ledger write. Records are validated and normalized on arrival.

request

POST /v1/records
Authorization: Bearer sk_live_••••••••
 
{
  "source": "internal_ledger",
  "records": [
    {
      "ref": "txn_9f21c",
      "amount": 184.50,
      "currency": "EUR",
      "settled_at": "2026-07-09T14:02:00Z"
    }
  ]
}

response

200 OK
 
{
  "accepted": 1,
  "normalized": [
    {
      "id": "rec_82ad1",
      "ref": "txn_9f21c",
      "schema": "ledger.v1",
      "status": "reconciling"
    }
  ]
}

Consume

Every result, structured and queryable.

Differences, AI-proposed resolutions, tickets, and side-by-side comparisons are all available as plain JSON — build your own dashboard, bot, or admin tool on top.

differences.json

GET /v1/differences?status=open
 
{
  "differences": [
    {
      "id": "diff_3b7e0",
      "check": "sepa_credit",
      "delta": -2847.00,
      "currency": "EUR",
      "suggested_cause": "duplicate_file_rerun",
      "confidence": 0.92,
      "ticket_id": "tkt_20260708-11"
    }
  ],
  "next_cursor": "eyJpZCI6..."
}

API Reference

A small, predictable surface.

No SDK lock-in required — a handful of REST endpoints cover ingestion, querying, and workflow.

POST/v1/recordsSend raw records from any source — schema detected automatically

GET/v1/checksList every configured reconciliation check and its live match rate

GET/v1/differencesQuery unresolved breaks — filterable by check, age, and amount

GET/v1/differences/:id/resolutionFetch the AI-suggested cause, confidence, and proposed fix

POST/v1/ticketsOpen, assign, or close a ticket tied to a difference

GET/v1/checks/:id/comparisonSide-by-side totals and match rate for any check and period

POST/v1/webhooksRegister an endpoint for match, break, and resolution events

MCP

Also available: an MCP server for asking your data questions.

Alongside the API, we run an MCP server so an agent — yours or your team's — can look up differences, tickets, and resolutions in conversation. It's a companion for ad hoc lookups, not where you build your integration.

REST API

For your integration

Deterministic calls your code controls

Scheduled jobs, webhooks, and cron-driven pulls

Powers dashboards, bots, and automation at volume

Runs unattended, no LLM in the loop

MCP server

For looking things up

Ask an agent about open differences or a specific ticket

Trigger a one-off action mid-conversation

Useful for analysts, support, and exploring during dev

Not for scheduled jobs or high-volume automation

MCP calls the same API underneath — it isn't a separate backend or a replacement for it. If you're building the actual pipeline your ops tools run on, that's the REST API.

How it works

From raw record to reconciled data, over API.

Send

Push records from your systems the moment they exist — via API or event stream. No file exports, no scheduled uploads.

Reconcile

Records are normalized against a common schema and matched continuously. Every difference is scored and a likely cause attached.

Consume

Pull differences, resolutions, tickets, and comparisons back into your own tools — dashboards, Slack bots, or internal admin panels.

Why engineering-first

Same reconciliation. No UI required.

Operations gives an ops team a workspace to work inside. Engineering first gives your team the same matching engine as an API, so the workspace is whatever you already build internally.

Data stays yours

Consume results via API and store them wherever you already keep operational data. Nothing is locked behind our UI.

Built for your stack

REST API, typed SDKs, and webhooks — drop reconciliation into an existing service instead of asking a team to adopt a new tool.

Same engine, no UI required

The identical matching engine behind Operations runs headless here — you get the reconciliation, you build the interface.

Teams build things like

Internal ops dashboardsSlack / Teams alerting botsCustom exception queuesFinance close checklistsCustomer-facing status pagesData warehouse sync jobs