AI-Enhanced Maintenance Operations for SFR Portfolio

The operator's property management system was tightly coupled with accounting, limiting flexibility. Manual maintenance workflows — from intake through dispatch and close — led to misclassified work orders, unnecessary dispatches, pricing variance, and slow resolution times across a geographically dispersed SFR portfolio.

We implemented a modular, AI-enhanced maintenance workflow spanning four stages. Triage uses a multimodal LLM for automated issue classification, risk-based prioritization, and remote troubleshooting. Intake leverages photo and video analysis to auto-structure requests and detect duplicates. Dispatch optimizes vendor matching by price, rating, proximity, and skillset with dynamic routing and price benchmarking. Close automates invoice validation, before/after image verification, CapEx vs. OpEx classification, and vendor performance scoring.