Skip to main content
Back to challenges
Challenge

Mastra AI Casualty ILS Trigger Analysis & Modeling

Insurance-linked security managers require continuous monitoring of casualty loss developments to compute collateral unlock triggers. Build a Mastra AI workflow that parses loss development triangles and calculates casualty ILS tranche payout probabilities.

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.

Utilize Mastra AI automated workflows to aggregate actuarial claims data and evaluate casualty ILS triggers.

Delivery guide

How work is evaluated

Evaluation

Evaluates accuracy of projected ultimate losses and correctness of collateral lock/unlock triggers.

Datasets

Challenge datasets

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

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

collateral_action_check

Verifies correct collateral decision based on attachment threshold

Binary check

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

Dimension 2

ultimate_loss_mae

Mean absolute error on ultimate loss estimation • target: 100000 • range: 0-5000000

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

  • Build a Mastra AI workflow for dynamic casualty claims triangle aggregation

  • Model loss development factors (LDF) to project ultimate losses for casualty ILS tranches

  • Trigger automated collateral release notifications when loss thresholds remain breach-free

Resources and assets

Reference links and supporting material

Dataset notes

Actuarial loss development triangles for liability lines supporting casualty ILS deals.

Challenge setup

Evaluates accuracy of projected ultimate losses and correctness of collateral lock/unlock triggers.

Challenge input

JSON containing cumulative_loss_triangle and attachment_point

Mastra AI

Provides TypeScript workflow graphs for complex multi-stage financial calculations

Evaluated output

JSON with ultimate_loss_estimate, trigger_breached (bool), collateral_action

Evaluation checks
  • Verifies correct collateral decision based on attachment threshold
  • Mean absolute error on ultimate loss estimation • target: 100000 • range: 0-5000000
Configured evaluation evidence
  • Benchmark: ILS-Trigger-Bench
  • Ultimate Loss Mae target: 100000
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript 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

Action Space
  • Mastra AI · Required
  • Amazon Bedrock · Optional
  • llama.cpp · Optional
Orchestration
  • Mastra AI · Required

Evaluation contract

  • collateral_action_check · Weight 1
  • ultimate_loss_mae · Weight 1

Metrics

  • ultimate_loss_mae · Target 100000

Benchmarks

  • ILS-Trigger-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing cumulative_loss_triangle and attachment_point

    Output: JSON with ultimate_loss_estimate, trigger_breached (bool), collateral_action

    Execution harness: JavaScript

Evaluation tools

  • Mastra AI · Agent Workflow Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Mastra AI Casualty ILS Trigger Analysis & Modeling