Skip to main content
Back to challenges
Challenge

Fintech Market Analysis Swarm with AutoGen and BoTorch

Utilize AutoGen to build a multi-agent team capable of analyzing global fintech funding trends. This team will consist of a Data Analyst agent, a Market Researcher agent, and an Optimization agent powered by BoTorch. The goal is to ingest high-level funding data and optimize resource allocation predictions for venture firms. By using the GPT-5.3-Codex model as the primary supervisor, the system can perform advanced reasoning over sparse financial datasets. Ellipsis is integrated into the workflow to facilitate natural language chat interaction, allowing stakeholders to query the agent swarm on funding shifts and deal count fluctuations while utilizing Cleanlab to ensure the integrity and quality of the ingested market data.

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.

Create a multi-agent fintech analysis swarm using AutoGen, BoTorch for optimization, and GPT-5.3-Codex for supervisor reasoning.

Delivery guide

How work is evaluated

Evaluation

Assess the swarm's ability to produce consistent market insights.

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

Data Consistency

Check if Cleanlab catches synthetic outlier data

Binary check

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

Dimension 2

Reasoning Depth

Evaluate supervisor agent's explanation quality • target: 8 • range: 0-10

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

  • Build AutoGen multi-agent architectures with task-specific roles

  • Configure GPT-5.3-Codex model reasoning to handle complex market trend synthesis

  • Master the use of BoTorch for optimizing financial decision variables

  • Implement Cleanlab to detect and sanitize anomalous data points during ingestion

  • Develop conversational interfaces with Ellipsis for real-time agent feedback loops

Resources and assets

Reference links and supporting material

Dataset notes

Sample data for 1 tasks

Challenge setup

Assess the swarm's ability to produce consistent market insights.

Challenge input

JSON list of quarterly funding records

BoTorch

AI Workflow Automation · Workflow Runners

AutoGen

Agent Systems · Multi-Agent Systems

Ellipsis

AI debugging assistant

Evaluated output

Trend analysis summary

Evaluation checks
  • Check if Cleanlab catches synthetic outlier data
  • Evaluate supervisor agent's explanation quality • target: 8 • range: 0-10
Configured evaluation evidence
  • Reasoning Depth target: 8
  • 1 public reference case
Configured runtime
  • Python execution harness

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
  • BoTorch · Required
  • AutoGen · Optional
  • Ellipsis · Optional

Evaluation contract

  • Data Consistency · Weight 1
  • Reasoning Depth · Weight 1

Metrics

  • Reasoning Depth · Target 8

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON list of quarterly funding records

    Output: Trend analysis summary

    Execution harness: Python

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Fintech Market Analysis Swarm with AutoGen and BoTorch