Bridging infrastructure, data, and intelligence to build scalable digital systems. From telecom networks to LLM pipelines — I engineer systems that actually ship.
I started in Telecom Engineering — supervising microwave deployments, managing field teams across Pakistan's North Region, and learning firsthand how infrastructure actually fails (and how to keep it from doing so).
That systems mindset carried me into Project Management at Blutech Consulting, where I coordinated digital transformation for HBL, UBL, and RawBank — building Power BI dashboards, automating incident pipelines, and turning chaos into KPIs.
Now I'm building in AI and data engineering — constructing RAG pipelines, agentic workflows, and LLM-powered products. Not as a pivot, but as a natural extension of understanding systems end-to-end.
I understand both the strategy layer (why we build) and the systems layer (how it runs). That intersection is where I do my best work.
Across four domains — from neural networks to network infrastructure.
Roles, results, and the tools that got us there.
Leading cross-functional execution for an AI-driven platform delivering automation and data insights to enterprise clients. Bridging product, engineering, and operations with Agile ceremonies and structured delivery governance.
Coordinated digital transformation for HBL, UBL, and RawBank across finance, telecom, and retail. Built Power BI dashboards, automated incident pipelines, and managed multi-system data integration workflows. Automated Start-of-Shift production reporting via Python and Bash.
Led delivery of 3 large-scale telecom deployment projects (microwave, wireless, power systems). Supervised cross-functional teams of 12+ engineers across Pakistan's North Region, working with Nokia, Ericsson, and ZTE infrastructure.
Problems first, technology second.
Full RAG pipeline with LangGraph agent orchestration — domain-aware Q&A over PDFs and web URLs. Supports engineering, medical, and legal domains via Streamlit frontend.
Flask web app using Google Gemini to extract skills and experience from multiple PDFs, then ranks candidates against a JD using FAISS vector similarity. Dual UI: web dashboard + Streamlit.
Stateful conversational AI integrating OpenAI, Anthropic Claude, and Gemini via OpenRouter. Designed for turn-by-turn memory, context retention, and clean API abstraction across all three providers.
Autonomous AI agent built with Gemini LLM and LangGraph for multi-step agentic workflows — demonstrating practical experience with foundation models and agent design patterns.
Practical, scalable, and designed to outlast the demo.
Open to product roles, AI engineering opportunities, and interesting problems in fintech, SaaS, and enterprise AI.