Kairos: Open-source, real-time fraud and risk decisioning

Finic released Kairos, an open-source real-time fraud and risk decisioning platform positioned as an alternative to Sardine, Sift, Riskified and SEON, with TypeScript (@finic-ai/kairos) and Python (finic-kairos) SDKs, a CLI, an HTTP API and Docker self-hosting. Kairos pairs Finic's decision model Grayson, which answers plain-language questions about each request with calibrated probabilities and works with few labels, with an MCP server that lets Claude Code, Cursor, Codex or any agent build workflows and propose model variants that Kairos trains and tests on the customer's data, reporting ROC-AUC, PR-AUC, gain over baseline with a 95% interval, and precision and recall at thresholds. In the vendor's example, an agent designed six variants for shopify_order and shipped a logistic regression combining order numbers with Grayson's answers that scored ROC-AUC 0.949 on validation and 0.930 on test as version v6.

Kairos https://kairos.finic.ai decides every sign-up, login, order, payment and payout as it happens, and keeps each decision with the signals behind it. It is an open-source alternative to Sardine, Sift, Riskified and SEON, with two things they don't have: your coding agent tunes it , and decision models catch fraud before you have the labels to train on. - Real-time decisions https://github.com/finic-ai/kairos/blob/main/docs/concepts/decisions.md : one call returns approve, step up, review or block, with a risk score and the reasons, from the subject's history, the devices, cards and addresses it shares with others, and the session. - Workflows https://github.com/finic-ai/kairos/blob/main/docs/platform/workflows.md : ask a decision model, score with a model, branch, call webhooks, open cases, and decide again later. - Queues and cases https://github.com/finic-ai/kairos/blob/main/docs/case-management/queues.md : review work for people, whose resolutions become labels. - MCP server https://github.com/finic-ai/kairos/blob/main/docs/reference/mcp.md : Claude Code, Cursor or Codex read your decisions and build workflows, with every action recorded under its key. - Experiments https://github.com/finic-ai/kairos/blob/main/docs/platform/models.md : your agent proposes model variants, Kairos trains and tests them on your data, and the agent ships the winner. - Grayson https://finic.ai : Finic's decision model answers questions about each request in plain language. It works with few labels and reads unstructured data. - Models https://github.com/finic-ai/kairos/blob/main/docs/platform/models.md : logistic regression and gradient-boosted trees XGBoost-compatible on your numbers and Grayson's answers, trained in Kairos or uploaded. - SDKs https://github.com/finic-ai/kairos/blob/main/docs/reference/clients/typescript.md for TypeScript @finic-ai/kairos and Python finic-kairos , a CLI https://github.com/finic-ai/kairos/blob/main/docs/reference/clients/cli.md , an HTTP API https://github.com/finic-ai/kairos/blob/main/docs/reference/api.md , and self-hosting https://github.com/finic-ai/kairos/blob/main/docs/deployment/self-hosting.md with Docker. You send Kairos what happens in your product: sign-ups, logins, orders, payouts, chargebacks. It decides each request approve, step up, review or block and keeps every decision with the signals behind it. What makes Kairos different is who tunes it. Connect Claude Code, Cursor, Codex or any agent over MCP , and the agent can do the work of a fraud ML engineer: 1. Build the workflow. It defines what to decide, which questions to ask a decision model, and which fields to score. 2. Design experiments. It proposes variants: a decision model alone, logistic regression on your numbers, gradient-boosted trees, a stack of them, or the decision model's answers combined with your numbers in one model. 3. Let Kairos run them. The server trains each variant on your oldest labeled data, chooses on the next, and tests on the newest, then reports ROC-AUC, PR-AUC, the gain over a baseline with its 95% interval, and precision and recall at your thresholds. 4. Pick the winner and ship it. The agent explains its choice, applies the variant as the workflow's next version, and comes back when new labels arrive. You: "Our chargebacks went up. Find a better model for shopify order and publish it if it wins." Agent: designs six variants → create workflow experiment → reads the results → "Logistic regression on the order's numbers plus Grayson's answers wins: ROC-AUC 0.949 on validation, 0.930 on test. Publishing it as v6." Kairos is built around Grayson https://finic.ai , Finic's decision model. You ask it questions in plain language "Does the shipping address look like a reshipper?", "Is this email throwaway?" and it answers with calibrated probabilities. Decision models are strong where traditional ML is weak: - Few labels. Grayson works on day one, before you have the thousands of chargebacks a model needs to learn from. - Unstructured data. It reads descriptions, emails, names, addresses, notes and item titles, which a gradient-boosted tree can't. - Explanations. Every answer is a question you wrote, so every decision says why. Traditional models are strong where decision models are weak: they learn your own outcomes precisely, from your own numbers, and they cost nothing per decision. Kairos lets you combine both: Grayson's answers become features next to your fields in a logistic regression or gradient-boosted trees XGBoost-compatible, trained in Kairos itself , and experiments tell you which combination wins on your data. Sign up at kairos.finic.ai https://kairos.finic.ai . Create a sandbox environment, a secret key, and you're ready to send events. In the console, open Settings › MCP , turn on the tool groups your agent needs, and create a key. Then, in Claude Code: claude mcp add --transport http kairos https://kairos.finic.ai/mcp --header "Authorization: Bearer ksk …" Cursor, Codex and VS Code work the same way: see MCP for agents https://kairos.finic.ai/docs/reference/mcp . Then ask your agent to look at your decisions, build a workflow, or run an experiment. npm install @finic-ai/kairos or: pip install finic-kairos js import { Kairos } from "@finic-ai/kairos"; const kairos = new Kairos { apiKey: process.env.KAIROS API KEY , baseUrl: "https://kairos.finic.ai" } ; await kairos.defineAction "withdrawal", { subject: "user", hold: "Should this withdrawal be held because the account shows signs of fraud or takeover?", } ; kairos.track "login", { userId: "user 42", deviceId: "dev 9", ipCountry: "GB" } ; const decision = await kairos.decide { type: "withdrawal", userId: "user 42", usd: 900 } ; // { decision: "approve" | "step up" | "review" | "block", riskScore, reasons, … } You need Node.js 22, pnpm 10 and a Grayson API key from portal.finic.ai https://portal.finic.ai/keys . No database to set up: the server starts its own Postgres. pnpm install cp .env.example .env set GRAYSON API KEY pnpm kairos dev the API on http://localhost:8787 and the console on http://localhost:5273 For your team, docker compose up --build . See Self-hosting https://github.com/finic-ai/kairos/blob/main/docs/deployment/self-hosting.md and Local setup https://github.com/finic-ai/kairos/blob/main/docs/development/local-setup.md . The CLI pnpm kairos signup the local server, a new account pnpm kairos login --profile prod --url https://kairos.finic.ai --email you@example.com pnpm kairos keys create "My server" a secret key for your code pnpm kairos actions define withdrawal --data @withdrawal.json pnpm kairos import events transactions.csv --type transaction --ts created at --entity user=customer pnpm kairos decide withdrawal -e user=user 42 -p usd=900 pnpm kairos workflows run review payout -i user=user 42 --wait pnpm kairos --help every command To have kairos on your path, run pnpm link --global in apps/cli . - Shopify: decide every order, hold risky ones for review, and learn from chargebacks. To get the Kairos app for your store, reach the Finic team on Discord https://discord.gg/4zRwRCK6Mx or at support@finic.ai mailto:support@finic.ai . - Inbound webhooks: take events from providers such as Stripe. See Inbound webhooks https://github.com/finic-ai/kairos/blob/main/docs/platform/webhooks/inbound.md . The docs are at kairos.finic.ai/docs https://kairos.finic.ai/docs , and in docs/ https://github.com/finic-ai/kairos/blob/main/docs/README.md . | Overview https://github.com/finic-ai/kairos/blob/main/docs/getting-started/overview.md | What Kairos does, how a decision is made, and what it does not do | | Quickstart https://github.com/finic-ai/kairos/blob/main/docs/getting-started/quickstart.md | A first decision in a few minutes | | MCP for agents https://github.com/finic-ai/kairos/blob/main/docs/reference/mcp.md | Connect Claude Code, Cursor, Codex or VS Code | | Models and experiments https://github.com/finic-ai/kairos/blob/main/docs/platform/models.md | Scoring models, and experiments that compare them | | Workflows https://github.com/finic-ai/kairos/blob/main/docs/platform/workflows.md | Triggers, steps, expressions and runs | | Concepts https://github.com/finic-ai/kairos/blob/main/docs/concepts/core-concepts.md | Entities, events, labels, links, actions, policies, decisions | | Import your history https://github.com/finic-ai/kairos/blob/main/docs/getting-started/import-history.md | Load existing events, customers and outcomes from files | | TypeScript SDK https://github.com/finic-ai/kairos/blob/main/docs/reference/clients/typescript.md · Python SDK https://github.com/finic-ai/kairos/blob/main/docs/reference/clients/python.md | The two clients | | API reference https://github.com/finic-ai/kairos/blob/main/docs/reference/api.md | Every HTTP endpoint | | Queues and cases https://github.com/finic-ai/kairos/blob/main/docs/case-management/queues.md | Review work for people | | Self-hosting https://github.com/finic-ai/kairos/blob/main/docs/deployment/self-hosting.md | Docker, Postgres, migrations, configuration, security | | Working on Kairos https://github.com/finic-ai/kairos/blob/main/docs/development/development.md | The repository, the server, the tests and the evaluation suite | | Path | What it is | |---|---| | apps/server | The HTTP API and MCP server. Hono on Node.js, Postgres through Drizzle. | | apps/console | The web console. React and Vite. | | apps/cli | kairos , the CLI: starts Kairos, connects to a server, and calls its API. | | packages/core | The decision path, the model scorers and the tree trainer. | | packages/sdk | @finic-ai/kairos , the TypeScript client. | | packages/sdk-python | finic-kairos , the Python client. | | packages/eval | The evaluation suite and the experiments. | | templates/ | Action definitions and event shapes to copy. | | examples/ | Runnable scenarios, end to end. | | docs/ | The documentation. | pnpm test all tests; no network pnpm test:python the tests of the Python SDK pnpm typecheck - Discord: questions, ideas and show-and-tell, at discord.gg/4zRwRCK6Mx https://discord.gg/4zRwRCK6Mx - Email: support@finic.ai mailto:support@finic.ai - Contributing: issues and pull requests are welcome. Read Working on Kairos https://github.com/finic-ai/kairos/blob/main/docs/development/development.md first. Kairos is licensed under the Apache License 2.0 https://github.com/finic-ai/kairos/blob/main/LICENSE .