work-orders

An end-to-end data engineering pipeline project that processes and analyzes Maintenance Work Orders using Mage, Docker, Google BigQuery, MariaDB, and Looker Studio. It features a seamless integration of cloud and open-source tools for scalable data storage, transformation, and visualization.

SmartGridBench: a smart-grid transformer-maintenance benchmark for tool-using LLM agents, built on IBM AssetOpsBench with MCP/direct-call tool surfaces.

Field techs text photos from a job site → AI-triaged work orders in Airtable, one job per trade with subtasks per unit, confirmed by SMS. n8n + two-model AI (vision + brain) + Twilio.

Track model cost while turning a field visit into a dispatcher note.

Self-hosted service management for small MSPs — managed clients and recurring billing, plus break/fix ticketing and work orders

Route field-service work-order follow-ups through an OpenAI-compatible model endpoint.

SmartGridBench: a smart-grid transformer-maintenance benchmark for tool-using LLM agents, built on IBM AssetOpsBench with MCP/direct-call tool surfaces.

Schedule a periodic field-service photo cleanup after dispatch and technician follow-up are complete.

Operations management platform built with Django 5.2 and DRF: work orders with a validated status machine, staff assignment, recurring task scheduler, RBAC, immutable audit trail, notifications, reports, and a documented JWT REST API.

Report field-service photos and dispatch follow-up metrics from a Python work-order workflow.

Track work-order photo failures and technician follow-up in a field-service agent loop.

Route field-service work-order summaries through an OpenAI-compatible gateway.

Python field-service work-order CSV export with a signed download URL.

Fan out field-service work-order updates to dispatch, customers, and technician leads.