AI Service Layer for Property Management Platform

Sole developer of the Python/FastAPI AI service layer for a commercial property management platform, integrated with a Java Spring Boot backend and Angular frontend across a distributed team. • AI anomaly detection and recommendation engine over property financial data, with trend-aware severity escalation and annualized cost-impact projection • Financial dashboard API endpoints covering NOI, OpEx ratio, P&L by category and seasonal baseline modelling • Receipt and invoice OCR pipeline using a vision LLM, with multilingual expense categorization • AI weather-risk alerting combining forecast data with historical climate normals • Vendor performance analysis using sentiment analysis over free-text tenant reviews • Automated PDF reporting (ReportLab, matplotlib) for quarterly and semi-annual investor reports • Tenant renewal scoring with deterministic decision gating and LLM output guards Stack: Python, FastAPI, PostgreSQL, Groq/LLaMA, Pandas, ReportLab, matplotlib, Git