AI Automation Engineer for Agentic Property Maintenance Platform

We’re building an internal system to automate how maintenance requests are handled across a portfolio of rental properties. Tenant reports an issue, and someone manually reviews it, checks the property history, contacts a contractor, follows up, and updates the record. We want an AI-driven workflow to handle most of this process. What the system should do When a request arrives by email or web form, the system should: Understand and categorize the issue Extract tenant, property, urgency, and problem details Check property/maintenance history Decide which workflow should run Contact the appropriate contractor through email/API Track responses and automatically follow up Escalate cases requiring human approval Update the maintenance record Provide a clear status summary The important part is that this should be a stateful agentic workflow, not just an LLM chatbot. The agent needs to use tools, make decisions, maintain case state, and move the request through multiple stages. Technical direction: Experience with the following is preferred: Agentic workflows / tool calling LangGraph, Temporal, or similar orchestration LLM APIs PostgreSQL Email/API integrations Background jobs and scheduled tasks Human-in-the-loop workflows For the MVP, we’ll focus on a few common maintenance scenarios and build the workflow end-to-end before expanding it. When applying, share one real agentic automation system you’ve built and briefly explain what decisions/actions the agent was responsible for.