Institutional Sale-Leaseback Underwriting Engine with Pydantic AI
Sale-leaseback transactions are becoming dominant in corporate real estate monetization, requiring tight validation of lease term inputs and exit cap rates. Build a type-safe underwriting agent using Pydantic AI to validate corporate financial disclosures and compute unlevered buyer internal rates of return (IRR). Achieve zero runtime schema errors and precise financial outputs.
What you are building
The core problem, expected build, and operating context for this challenge.
Build a type-safe sale-leaseback underwriting engine using Pydantic AI to compute buyer IRRs and validate tenant lease creditworthiness.
How work is evaluated
Evaluates accuracy of calculated initial cap rate and 10-year unlevered IRR for sale-leaseback investments.
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.
cap_rate_exact
Verifies exact calculation of initial cap rate.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
irr_mae
Mean Absolute Error in unlevered 10-year IRR output. • target: 0.002 • range: 0-0.05
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Implement Pydantic AI Agent models with strict schema validation logic
Define Pydantic models for corporate income statements and sale-leaseback lease structure
Compute initial cap rate, net rent escalations, and 10-year unlevered target IRR
Handle validation retries automatically when model inputs violate schema boundaries
Reference links and supporting material
Dataset of corporate sale-leaseback property offerings including initial purchase prices, starting rent, escalators, and exit cap assumptions.
Challenge setup
Evaluates accuracy of calculated initial cap rate and 10-year unlevered IRR for sale-leaseback investments.
Challenge input
JSON object with purchase_price, initial_annual_rent, annual_escalation_pct, exit_cap_rate
Pydantic AI
Mandatory assigned framework.
Evaluated output
JSON object with initial_cap_rate, unlevered_irr_10yr
- Verifies exact calculation of initial cap rate.
- Mean Absolute Error in unlevered 10-year IRR output. • target: 0.002 • range: 0-0.05
- Benchmark: FinancialUnderwritingBench
- Irr Mae target: 0.002
- 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
- Ellipsis · Optional
- Vercel AI Gateway · Optional
- Pydantic AI · Required
- Vercel AI Gateway · Optional
Evaluation contract
- cap_rate_exact · Weight 1
- irr_mae · Weight 1
Metrics
- irr_mae · Target 0.002
Benchmarks
- FinancialUnderwritingBench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON object with purchase_price, initial_annual_rent, annual_escalation_pct, exit_cap_rate
Output: JSON object with initial_cap_rate, unlevered_irr_10yr
Execution harness: Python
Evaluation tools
- Pydantic AI · Type-Safe Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.