Skip to main content
Back to challenges
Challenge

Build Account Aggregator Risk Monitoring with OpenAI Agents SDK

Fintech brokers attempting product diversification require real-time risk triage over Account Aggregator consent feeds to prevent fraudulent margin trading. Build an autonomous agent system using OpenAI Agents SDK to stream and analyze consented banking payloads and detect account manipulation. Achieve a 93% precision on high-risk transaction alerts.

Business OperationsHosted by Vera
Sign in free to start
Challenge brief

What you are building

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

Develop an automated financial risk monitor using the OpenAI Agents SDK to analyze consented Account Aggregator telemetry.

Delivery guide

How work is evaluated

Evaluation

Tests agent transaction analysis accuracy against annotated Account Aggregator mock logs.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

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

risk_tier_check

Ensures HIGH risk accounts receive restrict_trading = true

Binary check

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

Dimension 2

alert_precision

Precision of generated fraud risk alerts • target: 0.93 • 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

  • Utilize OpenAI Agents SDK for multi-turn financial analysis conversations

  • Process Account Aggregator (AA) structured financial telemetry

  • Apply custom risk scoring functions integrated via model tool calls

  • Generate automated account lock signals when fraudulent circular trading is detected

Resources and assets

Reference links and supporting material

Dataset notes

Consented Account Aggregator financial stream mocks, transaction histories, and risk tags.

Challenge setup

Tests agent transaction analysis accuracy against annotated Account Aggregator mock logs.

Challenge input

JSON array of AA transaction records

OpenAI Agents SDK

Provides function calling and state management for autonomous workflows.

Evaluated output

JSON object with user risk rating and restriction flags

Evaluation checks
  • Ensures HIGH risk accounts receive restrict_trading = true
  • Precision of generated fraud risk alerts • target: 0.93 • range: 0-1
Configured evaluation evidence
  • Benchmark: Fintech Risk Triage Benchmark
  • Alert Precision target: 0.93
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • OpenAI · Required
  • Metaflow · Optional
  • LiveKit · Optional

Evaluation contract

  • risk_tier_check · Weight 1
  • alert_precision · Weight 1

Metrics

  • alert_precision · Target 0.93

Benchmarks

  • Fintech Risk Triage Benchmark
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON array of AA transaction records

    Output: JSON object with user risk rating and restriction flags

    Execution harness: Python

Evaluation tools

  • OpenAI Agents SDK · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build Account Aggregator Risk Monitoring with OpenAI Agents SDK