Heatwave Microclimate Parametric Audit Agent using Pydantic AI and Weather Telemetry
During El Niño cycles, urban heat islands experience extreme temperature duration spikes that trigger parametric heat stress insurance policies for outdoor laborers and municipality electrical grids. Auditing millions of microclimate weather station readings requires type-safe Python agent workflows that guarantee zero hallucination in mathematical threshold validation. In this challenge, you will design a type-safe microclimate audit agent using Pydantic AI. The agent will ingest real-time weather API feeds, validate sensor calibration histories, verify wet-bulb globe temperature (WBGT) thresholds, and trigger automated micro-payouts. Your solution will execute dependency injection of weather database connections, apply Pydantic data validation, and emit fully auditable parametric settlement certificates.
What you are building
The core problem, expected build, and operating context for this challenge.
Build a type-safe heatwave audit agent with Pydantic AI to validate microclimate WBGT telemetry and trigger parametric payouts.
How work is evaluated
Evaluates Pydantic AI agent output validity when verifying WBGT parametric heat wave threshold breaches.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
pydantic_type_correctness
Output is valid Pydantic model instance
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
trigger_decision_accuracy
Accuracy of parametric trigger evaluation • target: 1 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Construct a type-safe AI agent using Pydantic AI.
Integrate dependency injection patterns for database connections and weather APIs.
Calculate Wet-Bulb Globe Temperature (WBGT) heat index metrics from raw sensor feeds.
Automate verifiable parametric policy trigger decisions.
Reference links and supporting material
Synthetic time-series of urban heat sensor readings including dry bulb temperature, relative humidity, and solar radiation during El Niño heat events.
Challenge setup
Evaluates Pydantic AI agent output validity when verifying WBGT parametric heat wave threshold breaches.
Challenge input
JSON with city_id, daily_max_wbgt_c, policy_threshold_c
Pydantic AI
Mandatory framework for type-safe agent execution.
Evaluated output
JSON matching HeatwaveAuditResult (triggered, consecutive_days, payout_usd)
- Output is valid Pydantic model instance
- Accuracy of parametric trigger evaluation • target: 1 • range: 0-1
- Benchmark: MicroclimateTriggerBench
- Trigger Decision Accuracy target: 1
- 1 public reference case
- 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
- Pydantic AI · Required
- Llama 3.1 405B · Optional
- Vast.ai · Optional
- Pydantic AI · Required
Evaluation contract
- pydantic_type_correctness · Weight 1
- trigger_decision_accuracy · Weight 1
Metrics
- trigger_decision_accuracy · Target 1
Benchmarks
- MicroclimateTriggerBench
- security
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with city_id, daily_max_wbgt_c, policy_threshold_c
Output: JSON matching HeatwaveAuditResult (triggered, consecutive_days, payout_usd)
Execution harness: Python
Evaluation tools
- Pydantic AI · Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.