Skip to main content
Back to challenges
Challenge

Build a CrewAI & OpenHands Epidemiological Outbreak Response Team

Public health agencies require automated multi-agent simulation tools to quickly model pathogen transmission dynamics and policy interventions. Build a CrewAI orchestrator that uses OpenHands code execution agents to run compartmental SIR models and generate outbreak mitigation strategies.

Data ScienceHosted 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 CrewAI agents collaborating with OpenHands code execution to model public health epidemiological outbreaks.

Delivery guide

How work is evaluated

Evaluation

Evaluates CrewAI crew performance in generating epidemiological simulations and policy outputs.

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

effective_r_reduction

Ensures effective R value under intervention is strictly lower than base R0.

Binary check

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

Dimension 2

simulation_accuracy

Accuracy of differential equation solving relative to standard ODE solutions. • target: 0.98 • 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

  • Design specialized agent roles (Epidemiologist, Data Scientist, Policy Advisor) in CrewAI

  • Leverage OpenHands agentic sandbox execution to write and run SciPy ODE epidemiological models

  • Compute R0 transmission metrics and simulated hospitalization curves under varying policy conditions

  • Synthesize clear public health operational directives supported by model predictions

Resources and assets

Reference links and supporting material

Dataset notes

Parameter set containing transmission constants, host population demographics, and intervention efficacy estimates for 15 outbreak scenarios.

Challenge setup

Evaluates CrewAI crew performance in generating epidemiological simulations and policy outputs.

Challenge input

JSON with r0_value (float), population (int), and policy_intervention (string).

CrewAI

Coordinates specialized role-playing sub-agents.

OpenHands

Executes programmatic modeling code in isolated sandboxes.

Evaluated output

JSON containing estimated_r_eff (float), peak_hospitalizations (int), and policy_advice (string).

Evaluation checks
  • Ensures effective R value under intervention is strictly lower than base R0.
  • Accuracy of differential equation solving relative to standard ODE solutions. • target: 0.98 • range: 0-1
Configured evaluation evidence
  • Benchmark: EpiModel Policy Benchmark
  • Simulation Accuracy target: 0.98
  • 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

Environment
  • OpenHands · Required
Action Space
  • OpenHands · Required
  • CrewAI · Optional
  • Qwen2.5-72B · Optional
Orchestration
  • OpenHands · Required

Evaluation contract

  • effective_r_reduction · Weight 1
  • simulation_accuracy · Weight 1

Metrics

  • simulation_accuracy · Target 0.98

Benchmarks

  • EpiModel Policy Benchmark
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with r0_value (float), population (int), and policy_intervention (string).

    Output: JSON containing estimated_r_eff (float), peak_hospitalizations (int), and policy_advice (string).

    Execution harness: Python

Evaluation tools

  • CrewAI · Agent Orchestrator
  • OpenHands · Code Execution System

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a CrewAI & OpenHands Epidemiological Outbreak Response Team