Skip to main content
Back to challenges
Challenge

Asset Management Fund Tax Incentive Validator with Pydantic AI & GPT-5 Pro

Singapore's asset management sector requires automated tax incentive eligibility calculations under updated MAS schemes. Build a type-safe qualification workflow using Pydantic AI and GPT-5 Pro to parse fund structures and output validated tax scheme recommendations.

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.

Construct a type-safe asset management tax eligibility agent using Pydantic AI and GPT-5 Pro reasoning models.

Delivery guide

How work is evaluated

Evaluation

Evaluates accuracy of structured fund eligibility output against MAS tax rules.

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

strict_pydantic_validation

Ensures schema strictness

Binary check

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

Dimension 2

eligibility_accuracy

Correct classification of eligibility status • 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

  • Implement structured dynamic models using Pydantic AI

  • Leverage GPT-5 Pro for complex legal taxonomy reasoning

  • Validate fund spend and local employment minimum threshold requirements

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic asset fund profiles with varying AUM, staffing, and local expenditure figures.

Challenge setup

Evaluates accuracy of structured fund eligibility output against MAS tax rules.

Challenge input

JSON containing AUM, local spend, local professionals count

Pydantic AI

Type safety and schema validation

GPT-5 Pro

Complex regulatory reasoning

Evaluated output

JSON containing eligible_scheme, compliant, shortfall_details

Evaluation checks
  • Ensures schema strictness
  • Correct classification of eligibility status • target: 1 • range: 0-1
Configured evaluation evidence
  • Benchmark: TaxEval-Bench
  • Eligibility 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
  • Pydantic AI · Required
  • GPT-5 Pro · Optional
  • Cekura · Optional
Orchestration
  • Pydantic AI · Required

Evaluation contract

  • strict_pydantic_validation · Weight 1
  • eligibility_accuracy · Weight 1

Metrics

  • eligibility_accuracy · Target 1

Benchmarks

  • TaxEval-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing AUM, local spend, local professionals count

    Output: JSON containing eligible_scheme, compliant, shortfall_details

    Execution harness: Python

Evaluation tools

  • Pydantic AI · Structured Agent Framework
  • GPT-5 Pro · LLM Model

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Asset Management Fund Tax Incentive Validator with Pydantic AI & GPT-5 Pro