Skip to main content
Loading navigation
Back to challenges

Deploy MSME Credit Assessment Agents with Microsoft AutoGen

Evaluating micro and small enterprises (MSMEs) for lending requires cross-referencing GST filing compliance, banking velocity, and promoter disclosures. Build a multi-agent debate and consensus framework using Microsoft AutoGen where Risk Analyst and Business Evaluator agents perform peer assessment on credit applications. Achieve a 91% alignment with expert credit committee decisions.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct a multi-agent conversational consensus system using Microsoft AutoGen to evaluate MSME credit eligibility from GST and bank data.

How work is evaluated

Evaluation

Evaluates AutoGen multi-agent consensus decisions against human credit committee labels.

Datasets

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

How submissions are scored

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

consensus_validity

Ensures recommendation is explicitly APPROVE or REJECT

Binary check

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

Dimension 2

committee_alignment

Accuracy relative to expert credit committee benchmarks • target: 0.91 • 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

  • Configure multi-agent multi-turn group chats using Microsoft AutoGen

  • Establish distinct risk personas (Conservative Risk Officer vs Growth Evaluator)

  • Process GST tax filings alongside Account Aggregator bank cashflows

  • Generate unified, policy-compliant credit assessment reports through agent consensus

Resources

Dataset notes

GST turnover metrics, Account Aggregator monthly summaries, and loan performance history.

Challenge setup

Evaluates AutoGen multi-agent consensus decisions against human credit committee labels.

Challenge input

JSON object with GST turnover, AA cashflows, and loan request

AutoGen

Enables multi-agent conversational consensus and role-playing execution.

Evaluated output

JSON object with final consensus recommendation and approved limit

Evaluation checks
  • Ensures recommendation is explicitly APPROVE or REJECT
  • Accuracy relative to expert credit committee benchmarks • target: 0.91 • range: 0-1
Configured evaluation evidence
  • Benchmark: MSME Credit Consensus Benchmark
  • Committee Alignment target: 0.91
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • AutoGen · Required
  • Microsoft AutoGen · Optional
  • Aim · Optional

Evaluation contract

  • consensus_validity · Weight 1
  • committee_alignment · Weight 1

Metrics

  • committee_alignment · Target 0.91

Benchmarks

  • MSME Credit Consensus Benchmark
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with GST turnover, AA cashflows, and loan request

    Output: JSON object with final consensus recommendation and approved limit

    Execution harness: Python

Evaluation tools

  • AutoGen · Multi-Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Deploy MSME Credit Assessment Agents with Microsoft AutoGen