Distressed Commercial Asset Restructuring Engine with Google ADK & Claude 4 Opus
Commercial real estate distress requires sophisticated capital stack restructuring and debt-for-equity swap modeling to free up cash flow. Build a multi-tier financial restructuring simulator using Google ADK with Claude 4 Opus to model debt haircut scenarios and equity dilution structures. Achieve under 2% variance against institutional private equity modeling benchmarks.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop an agentic debt restructuring analysis tool using Google ADK powered by Claude 4 Opus for commercial real estate recovery.
How work is evaluated
Evaluates financial accuracy of post-swap interest coverage calculation and final equity ownership distribution.
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.
dscr_accuracy
Checks DSCR within tolerance.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
equity_pct_error
Absolute error on equity percentage allocation. • target: 0.01 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Initialize Google ADK (Agent Development Kit) agents targeting Claude 4 Opus
Model multi-tranche debt capital stacks (senior debt, mezzanine, preference equity)
Calculate post-restructuring Debt Service Coverage Ratios (DSCR) and equity value creation
Automate debt-for-equity term sheet negotiation reasoning
Reference links and supporting material
Dataset of distressed commercial real estate debt structures, interest obligations, operating NOIs, and restructuring scenarios.
Challenge setup
Evaluates financial accuracy of post-swap interest coverage calculation and final equity ownership distribution.
Challenge input
JSON object with senior_debt, interest_rate, noi, debt_converted_to_equity
Google ADK
Mandatory Google ADK framework requirement.
Claude 4 Opus
Mandatory LLM model for financial reasoning.
Evaluated output
JSON object with post_swap_dscr, lender_equity_pct
- Checks DSCR within tolerance.
- Absolute error on equity percentage allocation. • target: 0.01 • range: 0-1
- Benchmark: RealEstateCapitalBench
- Equity Pct Error target: 0.01
- 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
- Google ADK · Required
- Distyl · Optional
- Claude 4 Opus · Optional
Evaluation contract
- dscr_accuracy · Weight 1
- equity_pct_error · Weight 1
Metrics
- equity_pct_error · Target 0.01
Benchmarks
- RealEstateCapitalBench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON object with senior_debt, interest_rate, noi, debt_converted_to_equity
Output: JSON object with post_swap_dscr, lender_equity_pct
Execution harness: Python
Evaluation tools
- Google ADK · Agent Architecture
- Claude 4 Opus · Reasoning LLM Model
Recipe freeze marker
No recipe freeze time is recorded.