Python/Streamlit Developer — Code Hardening, UI Polish, and AI-Powered Detection Feature

Scope (approx. 20–25 hours total): 1) Code Hardening (~6–8 hrs) - Review and clean up the existing codebase - Improve error handling and edge case resilience around core calculation logic - Remove dead code / inconsistent patterns from earlier iterations - Ensure the app runs reliably without crashing during normal use 2) UI Polish (~5–7 hrs) - Upgrade the Streamlit interface to a clean, professional standard - Improve layout, spacing, and visual hierarchy - Clean presentation of data tables and dashboard views 3) New Feature: Upload + AI Detection (~8–10 hrs) - Build a file upload handler that scans an uploaded data file for 5 pre-defined exception scenarios (spec will be provided) - Build detection logic to flag each scenario in the uploaded data - Wire an API call (Anthropic Claude API) that returns a plain-language explanation and suggested fix for each detected issue - Build a simple confirm/dismiss UI step for the user to review the suggestion What I'll provide: - Current codebase - Written spec for the 5 exception scenarios (data pattern + expected detection + expected response) - Sample test data Process: I'll review the full build, give one consolidated round of feedback, you complete one revision round, then I confirm and we close out. Please flag early if you think any phase will run meaningfully over its estimated hours so we can discuss before continuing.