Back to evaluations
Draft evaluation
Multi-Agent Real Estate Underwriting Engine with AutoGen and Galileo — evaluation
Evaluates the precision of debt metrics calculation and memo completeness against benchmark deal sheets.
Evaluation type
task based
Challenge
Multi-Agent Real Estate Underwriting Engine with AutoGen and Galileo
Difficulty
Advanced
Rigor
Not declared
The author has not specified a rigor level.
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
1
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
Underwriting Execution Test
Runs the multi-agent pipeline on asset dossier JSONs and evaluates calculated metrics against target values.
Input format
JSON containing property financial statements, NOI history, and debt terms
Output format
JSON containing project_irr, target_dscr, net_equity_multiple, and risk_flags