
1
1
Pakistan
2+ years of experience
AI developer and full-stack engineer from Faisalabad, Pakistan. I build AI-powered products end to end, from LLM-integrated apps and semantic search with vector embeddings to n8n automation pipelines that remove manual work. Currently a Software Developer at CyberGen AI Pakistan. I work across the MERN stack, Next.js, Angular, and React Native/Flutter, and I’m here to hack, ship fast, and team up with other builders.

StreetFix closes the loop between citizens and city operations teams. A citizen photographs a pothole, broken streetlight, blocked drain or overflowing bin, adds an optional note and location, and submits it in seconds. The report is analysed by an open model served on AMD infrastructure. StreetFix extracts the issue category, visible evidence, safety risk, severity and the responsible department. It then compares the new report with nearby open cases of the same category. If another citizen already reported the same problem, StreetFix merges the two into one case and explains why, instead of creating a duplicate ticket. Operations staff get a priority queue ranked by a transparent score: severity, number of reports, safety risk and age. Every point of the score is explained on screen. Staff can triage, assign, request more information and resolve a case, and each action is recorded in a public timeline. Citizens track their case with its ID and can ask questions such as "What is happening with my report?" The assistant answers only from the official case record, and if something has not been recorded it says so instead of guessing. The model never makes the risky decisions. Priority scoring and the status workflow are deterministic code, and every AI result is labelled with its source and model, so the app never claims AMD inference that did not happen. Built with Node.js, Express and a lightweight web frontend. Goals: fewer duplicate tickets, faster triage of dangerous issues and clearer accountability.
18 Oct 2026