Back to challenges
Challenge

Climate Heat Stress Risk Assessment using Mastra AI and o3

Rising European temperatures require real-time risk assessment for agricultural and municipal property portfolios under extreme thermal stress. Build an automated risk scoring workflow using Mastra AI and o3 model reasoning to generate explainable heat vulnerability indexes. Achieve a Mean Absolute Error under 0.05 on risk scores.

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 automated climate vulnerability scoring service using Mastra AI TypeScript workflows paired with OpenAI o3 model reasoning.

Delivery guide

How work is evaluated

Evaluation

Evaluates property heat vulnerability index predictions against geospatial ground truth.

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

score_range_check

Vulnerability score must be between 0.0 and 1.0.

Binary check

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

Dimension 2

mae

Mean Absolute Error on vulnerability index score • target: 0.04 • range: 0-0.2

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 Mastra AI agent workflows in TypeScript with structured state memory

  • Integrate o3 model reasoning for complex geospatial climate vulnerability scoring

  • Process raster geospatial thermal maps and municipal vulnerability indices

  • Achieve MAE < 0.05 against ground-truth climate vulnerability metrics

Resources and assets

Reference links and supporting material

Dataset notes

Geospatial dataset of 2,000 European urban properties with microclimate features and historical thermal loss values.

How this agent runs

Evaluates property heat vulnerability index predictions against geospatial ground truth.

Challenge input

JSON object with latitude, longitude, building_type, urban_heat_island_index, and green_cover_pct

Mastra AI

Offers TypeScript-native workflow state engines and tool integrations.

o3

Provides step-by-step logical reasoning over complex geospatial risk relationships.

Evaluated output

JSON object with vulnerability_score (float 0-1) and primary_risk_factor (string)

Checks for
  • Vulnerability score must be between 0.0 and 1.0.
  • Mean Absolute Error on vulnerability index score • target: 0.04 • range: 0-0.2
Proof of success
  • Benchmark: ClimateVulnerabilityBench
  • Mae target: 0.04
  • 1 public reference case
Runtime evidence
  • JavaScript execution harness
  • javascript sandbox (unavailable on Versalist)
View technical recipe

Configured tools

Action Space
  • Mastra AI · Optional
  • Baseten · Optional
Policy Serving
  • o3 · Required
Orchestration
  • Mastra AI · Optional

Evaluation contract

  • score_range_check · Weight 1
  • mae · Weight 1

Recipe state

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

Frequently Asked Questions about Climate Heat Stress Risk Assessment using Mastra AI and o3