Skip to main content
Back to challenges
Challenge

Orchestrate Focal Therapy Decision Workflows with CrewAI & o3

Focal therapy for localized prostate cancer presents complex trade-offs between oncological control and quality of life. Deploy a multi-agent consensus system using CrewAI and OpenAI o3 to analyze patient MRI, biopsy, and genomic data to produce evidence-backed treatment consensus with 88% clinician agreement.

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.

Multi-agent clinical decision framework built with CrewAI and OpenAI o3 model reasoning.

Delivery guide

How work is evaluated

Evaluation

Evaluates multi-agent tumor board agreement and recommendation alignment with guidelines.

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

guideline_compliance_test

Verifies recommendation matches evidence-based clinical guidelines.

Binary check

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

Dimension 2

consensus_agreement_rate

Percentage agreement with ground truth tumor board decision. • target: 0.88 • 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

  • Configure multi-agent role-based teams using CrewAI framework.

  • Utilize OpenAI o3 advanced reasoning capabilities for multi-criteria trade-off evaluations.

  • Synthesize MRI PIRADS scores and pathology reports into actionable tumor board recommendations.

Resources and assets

Reference links and supporting material

Dataset notes

Dataset of 40 multi-parametric MRI and pathology patient profiles with consensus expert tumor board decisions.

Challenge setup

Evaluates multi-agent tumor board agreement and recommendation alignment with guidelines.

Challenge input

JSON patient clinical profile including Gleason score, lesion PIRADS, and age

CrewAI

Orchestrates role-playing agents.

o3

OpenAI reasoning model for complex medical trade-off evaluation.

Evaluated output

JSON consensus object with recommended_modality and rationale

Evaluation checks
  • Verifies recommendation matches evidence-based clinical guidelines.
  • Percentage agreement with ground truth tumor board decision. • target: 0.88 • range: 0-1
Configured evaluation evidence
  • Benchmark: OncoMDT-Bench
  • Consensus Agreement Rate target: 88%
  • 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
  • CrewAI · Optional
  • Zed · Optional
Policy Serving
  • o3 · Required

Evaluation contract

  • guideline_compliance_test · Weight 1
  • consensus_agreement_rate · Weight 1

Metrics

  • consensus_agreement_rate · Target 0.88

Benchmarks

  • OncoMDT-Bench
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON patient clinical profile including Gleason score, lesion PIRADS, and age

    Output: JSON consensus object with recommended_modality and rationale

    Execution harness: Python

Evaluation tools

  • CrewAI · agent_framework
  • o3 · llm

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Orchestrate Focal Therapy Decision Workflows with CrewAI & o3