Solo Project
2026
10 technologies
About This Project
FieldOps is a commercial field service management platform engineered for dispatchers, technicians, and operations teams to orchestrate work order lifecycles, prevent overlapping bookings, and automate complaint classification. Built with Next.js 16 (App Router), React 19, Express 5, and PostgreSQL via Prisma ORM, the system enforces a deterministic state machine (DRAFT → SCHEDULED → DISPATCHED → IN_PROGRESS → COMPLETED) where transitions require mandatory prerequisites - such as verified labor line items prior to job closure and explicit audit logs for holds or cancellations.
To prevent scheduling conflicts across field teams, the dispatch engine executes mathematical overlap collision checks (existingStart < newEnd && existingEnd > newStart) rejecting double-bookings with detailed conflict diagnostics. The platform's automated customer service pipeline features a dual-engine AI triage architecture: incoming service complaints stream to an LLM for structured trade categorization, urgency assessment, and technician competency matching, guarded by an automatic sub-5ms deterministic keyword heuristic matrix fallback with 3-second timeout protection so dispatch workflows never fail.
Tech Stack
Key Features
- Deterministic 5-stage work order state machine with mandatory transition prerequisites and audit logging
- Mathematical scheduling conflict engine preventing overlapping technician assignments across active time slots
- Dual-engine AI triage system with live LLM classification and sub-5ms deterministic heuristic fallback
- Line-item billing system with automated 8.25% sales tax calculation, cent-accurate rounding, and status locks
- Role-Based Access Control (Admin, Dispatcher, Field Technician) with quick demo persona switching
The Outcome
Deployed live across Vercel and Render at fieldops.adibdev.me (Backend API: fieldops-api.adibdev.me) with open-source code on GitHub. Validated with comprehensive Jest and Supertest test suites covering state machine transitions, overlap collision algorithms, automated sales tax calculations, and benchmark AI triage datasets.