Skip to main content
Back to challenges
Challenge

AutoGen Reinsurance Monte Carlo Negotiation Simulator

Reinsurance treaty negotiators need interactive multi-agent simulations to model rate renewals and retention structure options. Construct an AutoGen multi-agent simulation where buyer, broker, and reinsurer agents negotiate cat loss attachment points.

Machine LearningHosted by Vera
Sign in free to start
Challenge brief

What you are building

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

Build an AutoGen agent conversation framework simulating market equilibrium pricing during Monte Carlo treaty renewals.

Delivery guide

How work is evaluated

Evaluation

Evaluates negotiation convergence, agreement rate, and distance of agreed rate from market equilibrium.

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

rational_deal_check

Ensures deal is agreed only when buyer budget >= seller minimum requirement

Binary check

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

Dimension 2

equilibrium_distance

Absolute difference between agreed price and true market clearing price • target: 0.005 • range: 0-0.05

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

  • Construct a multi-agent negotiation game in AutoGen representing Cedent, Broker, and Reinsurer

  • Incorporate expected loss distribution constraints into each agent's policy utility function

  • Simulate Monte Carlo Rendez-vous contract negotiations to find market pricing equilibrium

Resources and assets

Reference links and supporting material

Dataset notes

Monte Carlo treaty negotiation scenarios specifying attachment points, expected losses, and rate floors.

Challenge setup

Evaluates negotiation convergence, agreement rate, and distance of agreed rate from market equilibrium.

Challenge input

JSON containing cedent_max_budget, reinsurer_min_rol, and expected_loss

AutoGen

Provides multi-agent group chat primitives suited for negotiation modeling

Evaluated output

JSON with deal_agreed (bool), final_rol, and total_negotiation_rounds

Evaluation checks
  • Ensures deal is agreed only when buyer budget >= seller minimum requirement
  • Absolute difference between agreed price and true market clearing price • target: 0.005 • range: 0-0.05
Configured evaluation evidence
  • Benchmark: MultiAgent-Negotiation-Bench
  • Equilibrium Distance target: 0.005
  • 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
  • AutoGen · Required
  • Anyscale · Optional
Policy Serving
  • GPT-4o mini · Optional

Evaluation contract

  • rational_deal_check · Weight 1
  • equilibrium_distance · Weight 1

Metrics

  • equilibrium_distance · Target 0.005

Benchmarks

  • MultiAgent-Negotiation-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing cedent_max_budget, reinsurer_min_rol, and expected_loss

    Output: JSON with deal_agreed (bool), final_rol, and total_negotiation_rounds

    Execution harness: Python

Evaluation tools

  • AutoGen · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about AutoGen Reinsurance Monte Carlo Negotiation Simulator