Skip to main content
Loading navigation
Back to challenges

Build an Actuarial Climate Risk Evaluation Engine with Mastra AI and Claude 4 Sonnet

Insurance actuaries need advanced climate scenario models to underwrite regional infrastructure risks accurately. Build a TypeScript workflow using Mastra AI and Claude 4 Sonnet to extract property exposure metrics, calculate climate loss projections, and format actuarial risk reports.

Workflow AutomationHosted by Vera
Sign in free to start

What you are building

Develop a TypeScript workflow using Mastra AI and Claude 4 Sonnet to perform structured climate risk assessment for insurance portfolios under MAS risk guidelines.

How work is evaluated

Evaluation

Evaluates Mastra AI workflow output quality against standard actuarial climate loss projection benchmarks.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

risk_tier_classification

Ensures property with elevation < 1.5m and high SLR scenario is classified HIGH risk

Binary check

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

Dimension 2

eal_accuracy

Calculated EAL within 5% tolerance of reference physics model • target: 0.95 • 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

  • Build TypeScript agentic workflows using Mastra AI framework

  • Leverage Claude 4 Sonnet for quantitative actuarial reasoning and analysis

  • Integrate geospatial climate flood hazard layers into underwriting steps

  • Generate MAS-compliant climate risk disclosure summaries

Resources

Dataset notes

Synthetic coastal asset inventory and climate vulnerability curves for Southeast Asian ports.

Challenge setup

Evaluates Mastra AI workflow output quality against standard actuarial climate loss projection benchmarks.

Challenge input

JSON object with asset_location, structural_value_usd, sea_level_rise_scenario_m

Mastra AI

Provides TypeScript workflow engine with state control

Claude 4 Sonnet

Delivers superior mathematical and technical document reasoning

Evaluated output

JSON object with risk_tier, expected_annual_loss_usd, and mas_compliance_summary

Evaluation checks
  • Ensures property with elevation < 1.5m and high SLR scenario is classified HIGH risk
  • Calculated EAL within 5% tolerance of reference physics model • target: 0.95 • range: 0-1
Configured evaluation evidence
  • Benchmark: ActuaryBench-2026
  • Eal Accuracy target: 0.95
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • Mastra AI · Required
Observation
  • Langfuse · Optional
Reward / Eval
  • Langfuse · Optional
Policy Serving
  • Claude 4 Sonnet · Optional
Orchestration
  • Mastra AI · Required

Evaluation contract

  • risk_tier_classification · Weight 1
  • eal_accuracy · Weight 1

Metrics

  • eal_accuracy · Target 0.95

Benchmarks

  • ActuaryBench-2026
  • materials

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with asset_location, structural_value_usd, sea_level_rise_scenario_m

    Output: JSON object with risk_tier, expected_annual_loss_usd, and mas_compliance_summary

    Execution harness: JavaScript

Evaluation tools

  • Mastra AI · Agent Workflow Framework
  • Claude 4 Sonnet · LLM Model

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build an Actuarial Climate Risk Evaluation Engine with Mastra AI and Claude 4 Sonnet