Skip to main content
Back to challenges
Challenge

Actuarial Parametric Insurance Engine using LlamaIndex and Sarvam AI

Construct an automated climate risk evaluation agent using LlamaIndex and Sarvam AI models to assess catastrophe losses for regional parametric insurance payouts. Parse geospatial satellite weather feeds and generate automated claims trigger assessments.

Workflow AutomationHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Build a LlamaIndex RAG and tool-use engine empowered by Sarvam AI multilingual and domain models to evaluate climate risk indices for insurance contracts.

Delivery guide

How work is evaluated

Evaluation

Validates parametric policy claim evaluations against ground truth environmental thresholds.

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

trigger_accuracy

Ensures parametric condition triggers correctly

Binary check

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

Dimension 2

payout_accuracy

Accuracy of financial payout calculation • target: 1 • 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

What you should walk away with

  • Build specialized RAG indexes over actuarial hazard tables using LlamaIndex VectorStoreIndex

  • Incorporate Sarvam AI domain-adapted inference models for regional Southeast Asian environmental data processing

  • Implement automated parametric payout decision trees based on real-time weather indices

  • Produce actuarial audit notes adhering to Asian Actuarial Conference standards

Resources and assets

Reference links and supporting material

Dataset notes

Historical weather event data combined with parametric insurance contract policy parameters.

Challenge setup

Validates parametric policy claim evaluations against ground truth environmental thresholds.

Challenge input

JSON object with policy_id, contract_thresholds, and observed_weather_data

LlamaIndex

Structures unstructured actuarial documents and policy terms.

Sarvam AI

Provides domain-tuned LLM execution for climate text processing.

Evaluated output

JSON object with trigger_activated (boolean), calculated_payout, and actuarial_summary

Evaluation checks
  • Ensures parametric condition triggers correctly
  • Accuracy of financial payout calculation • target: 1 • range: 0-1
Configured evaluation evidence
  • Benchmark: Actuarial Quant Evaluation Suite
  • Payout Accuracy target: 1
  • 1 public reference case
Configured runtime
  • 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

Action Space
  • Llama Index · Required
  • Sarvam AI · Optional
  • Exa · Optional

Evaluation contract

  • trigger_accuracy · Weight 1
  • payout_accuracy · Weight 1

Metrics

  • payout_accuracy · Target 1

Benchmarks

  • Actuarial Quant Evaluation Suite
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with policy_id, contract_thresholds, and observed_weather_data

    Output: JSON object with trigger_activated (boolean), calculated_payout, and actuarial_summary

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Data Agent & Indexing Engine
  • Sarvam AI · Inference Engine

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Actuarial Parametric Insurance Engine using LlamaIndex and Sarvam AI