OFFICE-PATROL is an AI-powered autonomous contract enforcement system inspired by the problem of companies losing money because vendor SLA obligations are buried in contracts and often missed. It reads vendor contracts, extracts SLA conditions, deadlines, penalties, and exceptions, and connects them with live Freshdesk tickets to detect potential breaches. A multi-agent system then verifies the breach, checks financial limits and security risks, and decides whether to act automatically or request human approval. When a breach is confirmed, the system can communicate with the vendor through an AI voice agent, capture their commitment, update the ticket, and initiate eligible penalty recovery. We built it using a multi-agent architecture with contract extraction, ticket mapping, breach detection, verification, and vendor communication components, orchestrated through LangGraph and integrated with Freshdesk. During development, we faced challenges around contract ambiguity, AI hallucinations, prompt injection, financial safety, and reliable integration between multiple systems. We are proud that OFFICE-PATROL goes beyond being a chatbot by creating an end-to-end workflow from understanding a contract to detecting a breach and taking action. Through the project, we learned that effective agentic AI requires strong governance, deterministic rules, security checks, structured outputs, and clear boundaries for autonomous actions. Going forward, we plan to support more enterprise platforms, improve multilingual voice communication, strengthen predictive SLA monitoring and security, and expand OFFICE-PATROL into a broader enterprise contract and obligation management system.
office patrol
OFFICE-PATROL uses AI agents to detect vendor SLA breaches, verify them, negotiate resolutions, and automatically recover lost money.
Updates
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.