Challenge

Gigaproject Leadership & Schedule Governance Multi-Agent Simulator in AutoGen

Leadership transitions in Saudi Arabia real estate gigaprojects often trigger project schedule reviews. Build an AutoGen multi-agent conversational simulator featuring Developer, Contractor, and Governance agents that negotiate project timeline adjustments and output an updated milestone roadmap.

Business OperationsHosted by Vera
Challenge brief

What you are building

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

Utilize Microsoft AutoGen multi-agent conversation frameworks to simulate real estate gigaproject executive governance and schedule renegotiations.

Delivery guide

How work is evaluated

Evaluation

Evaluates AutoGen agents' ability to reach consensus on milestone extensions under leadership replacement scenarios.

Datasets

Shared data for this challenge

Review public datasets and any private uploads tied to your build.

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

Consensus State Test

Ensures conversational agents successfully reach consensus.

Binary check

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

Dimension 2

Negotiation Variance

Difference between proposed extension and compromise deadline. • target: 4 • range: 0-12

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 Microsoft AutoGen ConversationalAgent patterns for corporate governance

  • Simulate executive negotiation between project CEO, EPC contractors, and auditors

  • Re-evaluate gigaproject delivery milestones following executive replacements

  • Generate structured milestone revision summaries with human-in-the-loop review steps

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic leadership transition records and master schedule delay metrics for Saudi Arabia real estate gigaprojects.

How this agent runs

Evaluates AutoGen agents' ability to reach consensus on milestone extensions under leadership replacement scenarios.

Challenge input

JSON object with original_deadline, requested_extension_months, and chief_executive_status.

AutoGen

Multi-agent conversational negotiation and governance simulation

Evaluated output

JSON object with agreed_extension_months, consensus_reached, and revised_deadline.

Checks for
  • Ensures conversational agents successfully reach consensus.
  • Difference between proposed extension and compromise deadline. • target: 4 • range: 0-12
Proof of success
  • Benchmark: GCC Real Estate Governance Benchmark
  • Negotiation Variance target: 4
  • 1 public reference case
Runtime evidence
  • Python execution harness
  • Python sandbox (unavailable on Versalist)
View technical recipe

Configured tools

Action Space
  • AutoGen · Required
  • Synthflow · Optional
Policy Serving
  • OpenRouter · Optional
Orchestration
  • OpenRouter · Optional

Evaluation contract

  • Consensus State Test · Weight 1
  • Negotiation Variance · Weight 1

Recipe state

This is a preview. The configuration can change before the evaluation recipe is locked.

Frequently Asked Questions about Gigaproject Leadership & Schedule Governance Multi-Agent Simulator in AutoGen