Back to challenges
Challenge

Build a COSMIC AML Risk Intelligence Agent with Vercel AI SDK and Galileo

Construct an automated Anti-Money Laundering (AML) risk scoring and intelligence agent to analyze cross-institutional transaction logs using Vercel AI SDK and Galileo. Implement real-time suspicious activity reporting with zero regulatory leakage.

Workflow AutomationHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Orchestrate multi-bank transaction graph inspection using Vercel AI SDK tools integrated with Galileo evaluation guardrails for MAS COSMIC compliance.

Delivery guide

How work is evaluated

Evaluation

Evaluates the agent's ability to identify transaction laundering rings, enforce guardrails via Galileo, and produce accurate STR reports.

Datasets

Shared data for this challenge

Review public datasets and any private uploads tied to your build.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

schema_conformance

Ensures output JSON matches target regulatory schema

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

precision

Precision of laundering ring identification • target: 0.95 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Master building streaming agent workflows using Vercel AI SDK tool calling

  • Integrate Galileo for continuous hallucination monitoring and guardrail telemetry

  • Implement graph analytics algorithms to detect cross-institutional money laundering rings

  • Synthesize MAS-compliant Suspicious Transaction Report (STR) summaries with verifiable provenance

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic multi-bank ledger containing layered transactions across DBS, OCBC, and UOB mock accounts.

How this agent runs

Evaluates the agent's ability to identify transaction laundering rings, enforce guardrails via Galileo, and produce accurate STR reports.

Challenge input

JSON array of cross-bank transaction logs and account profile data

AI SDK

Provides TypeScript tool streaming and structured output generation.

Galileo

Evaluates hallucination and risk output in real time.

Evaluated output

JSON object with risk_score, laundering_ring_detected (boolean), and structured str_report

Checks for
  • Ensures output JSON matches target regulatory schema
  • Precision of laundering ring identification • target: 0.95 • range: 0-1
Proof of success
  • Benchmark: Financial Crime Detection Benchmark (FinCrimeBench)
  • Precision target: 0.95
  • 1 public reference case
Runtime evidence
  • JavaScript execution harness
  • javascript sandbox (unavailable on Versalist)
View technical recipe

Configured tools

Action Space
  • Vercel AI SDK · Optional
  • Claude Code · Optional
Observation
  • Galileo · Required
  • Claude Code · Optional
Reward / Eval
  • Galileo · Required
Orchestration
  • Vercel AI SDK · Optional

Evaluation contract

  • schema_conformance · Weight 1
  • precision · Weight 1

Recipe state

This is a preview. The configuration can change before the evaluation recipe is locked.

Frequently Asked Questions about Build a COSMIC AML Risk Intelligence Agent with Vercel AI SDK and Galileo