Back to challenges
Challenge

Geospatial Wildfire & Flood Resilience Trigger Agent with OpenAI Agents SDK

Climate physical risk modeling requires real-time agentic evaluation of asset resilience investments against severe weather triggers. You will construct an autonomous resilience evaluation agent using the OpenAI Agents SDK to analyze satellite flood inundation maps and property wildfire mitigation features. The agent must yield an optimal resilience ROI allocation with an MAE under 2.5%.

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.

Develop an autonomous climate resilience agent using the OpenAI Agents SDK to optimize property flood and wildfire mitigation investments.

Delivery guide

How work is evaluated

Evaluation

Evaluates climate resilience optimization accuracy across 100 high-risk coastal and wildfire zone properties.

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

valid_loss_reduction

Expected loss reduction must be between 0% and 100%.

Binary check

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

Dimension 2

roi_mae

Mean Absolute Error of expected loss reduction percentage relative to ground truth physics model. • target: 2.5 • range: 0-10

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

  • Initialize OpenAI Agents SDK agent using GPT-4o with custom function calling capabilities.

  • Connect agent to geospatial vector processing tools for flood and wildfire hazard layers.

  • Implement multi-turn function calling to evaluate property resilience ROI.

  • Generate auditable resilience investment proposals for risk engineering teams.

Resources and assets

Reference links and supporting material

Dataset notes

Dataset of 150 residential and commercial property profiles with geospatial hazard scores and structural mitigation cost tables.

How this agent runs

Evaluates climate resilience optimization accuracy across 100 high-risk coastal and wildfire zone properties.

Challenge input

JSON containing property_id, flood_zone, defensible_space_feet, budget_usd

OpenAI Agents SDK

Provides function calling loops and multi-turn conversation tools for GPT-4o.

Evaluated output

JSON containing recommended_investments (ARRAY), expected_loss_reduction_percent, net_roi_multiplier

Checks for
  • Expected loss reduction must be between 0% and 100%.
  • Mean Absolute Error of expected loss reduction percentage relative to ground truth physics model. • target:...
Proof of success
  • Benchmark: ClimateResilienceBench-2026
  • Roi Mae target: 2.5
  • 1 public reference case
Runtime evidence
  • Python execution harness
  • Python sandbox (unavailable on Versalist)
View technical recipe

Configured tools

Action Space
  • OpenAI · Required
  • Milvus · Optional
Policy Serving
  • Qwen 3 · Optional

Evaluation contract

  • valid_loss_reduction · Weight 1
  • roi_mae · Weight 1

Recipe state

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

Frequently Asked Questions about Geospatial Wildfire & Flood Resilience Trigger Agent with OpenAI Agents SDK