The compliance API

built for real time

The compliance API

built for real time

Ship transaction monitoring rules, risk scores and case workflows through one robust REST API, with typed SDKs and signed webhooks.

Read the docs

Talk to an engineer

Node.js

Python

OpenAPI

Webhooks

MCP

Transaction monitoring rules, as code

Version, test and activate through the API

Write monitoring rules as simple, readable configuration across transactions, clients, wallets and accounts, with rich aggregations over any time window. Every change is versioned. Start from a ready-made typology template, test it against your own history, then go live. Every decision comes with a clear explanation, and Corsa can check a payment in real time and stop it before it settles.

Rules API docs

POST /v1/rules

{
  "name": "High-risk client withdrawal velocity",
  "conditions": {
    "all": [
      { "entity": "client", "property": "riskTier",
        "operator": "equal", "value": "HIGH" },
      {
        "entity": "transaction",
        "aggregationProperty": "amount",
        "aggregationOperator": "sum",
        "aggregationTimeType": "in_the_last",
        "aggregationTimeValue": 1,
        "aggregationTimePeriod": "days",
        "aggregationFilters": [
          { "property": "type", "operator": "equal",
            "value": "WITHDRAW" }
        ],
        "operator": "greaterThanInclusive",
        "value": 50000
      }
    ]
  },
  "actions": [
    { "type": "CREATE_ALERT", "config": { "priority": "HIGH" } },
    { "type": "HALT_TRANSACTION", "config": {} }
  ]
}

Corporate CRRDraft

crr-model.json

Risk rating models you can read, test and version

Define, simulate and deploy your CRR model

Your customer risk rating model becomes clear, structured configuration: factors, score bands and thresholds, plus hard overrides such as a sanctions match. Your team builds it in Corsa’s Dynamic Risk Assessment, sees exactly who would move between low, medium and high before anything changes, then rolls it out with a recorded reason. Every version is kept, every score explains what drove it, and your monitoring rules use the result straight away. Already run a model elsewhere? Send its scores to Corsa instead.

Dynamic Risk Assessment

Corsa inside Claude and ChatGPT

One MCP server for your data and your Corsa agents

Connect Corsa’s MCP server to Claude, ChatGPT, Cursor or your own agents, sign in with OAuth, and work in plain language: summarize an alert, screen a client for sanctions and PEPs, pull analytics or research a UBO, then act on it by adding case notes, assigning work and updating alerts. Point the connection at one of your custom Corsa agents and talk to it from the tools your team already uses, with answers grounded in your live Corsa data.

MCP server docs

Corsa · api.corsa.finance/v1/mcp

OAuth or API key · scoped to your workspace

Built for production from day one

Reliability, security and scale are part of the platform, not something your engineers have to build around it.

Never double-count a transaction

Safe retries are built in, so a dropped connection never turns into a duplicate payment or a missed alert.

Works with the IDs you already have

Send records with your own identifiers and Corsa keeps them in sync, updating what changed instead of creating duplicates.

Built for volume

Load whole portfolios and update large batches of alerts and cases at once, without scripting around the API.

Real-time events you can trust

Every change in Corsa can trigger your own systems instantly, through signed and verifiable notifications.

Predictable at scale

Clear usage limits and end-to-end request tracing make integrations easy to scale, debug and support.

Your data stays in region

Run in the US or the EU on the same API and tools, with data kept where your regulators expect it.

Forward-deployed engineers on your team

Corsa’s FDEs work alongside your compliance team to design, engineer and ship the automation your program needs, not a generic setup.

Talk to an engineer

Map your process

An FDE sits with your compliance, risk and operations teams to map how alerts, reviews and escalations really run today, and where the manual work is.

Engineer the solution

They connect your data and vendors, then build the rules, risk model, workflows and AI agents your program needs on Corsa’s platform and API.

Test and go live

Rules are backtested and models simulated on your own history before launch, so your team signs off on real numbers, not estimates.

Keep automating

After launch, the same engineers tune thresholds, cut false positives and automate the next process as your volumes and products grow.

Frequently asked questions

Which languages can we build with?

Corsa offers official SDKs for Node.js and Python, and any other language can use the REST API directly. Your engineers work in the stack they already know.

Can we manage monitoring rules in code?

Yes. Create, test and activate rules through the API or the no-code Rule Builder, start from ready-made typology templates, and keep a full version history of every change.

Can Corsa stop a transaction before it settles?

Yes. Corsa can check a payment in real time and hold it before it settles whenever one of your rules calls for it.

Can we build our own risk rating model?

Yes. Your team defines factors, score bands, thresholds and overrides in Corsa's Dynamic Risk Assessment, previews the impact on your portfolio, then rolls it out. Already have a model? Send its scores to Corsa instead.

Can we use Corsa from Claude or ChatGPT?

Yes. Connect Corsa's MCP server to Claude, ChatGPT, Cursor or your own agents. Ask about alerts, cases, clients and transactions, research a counterparty, update your work, or talk to one of your custom Corsa agents.

Will Corsa help us build our setup?

Yes. Forward-deployed engineers work with your compliance team to map your processes and build the integrations, rules, risk model, workflows and agents your program needs, and they stay involved after launch.

How do we get started?

Talk to an engineer and we will set up your workspace. Your admins then create API keys for your team and services from the Developers Hub.