Player 1 · Islamabad, Pakistan
SOLUTION ARCHITECT & AI ENGINEER
I design and ship AI products: RAG and LLM systems, AI agents, computer vision and edge AI. As founder of Aeyron Technologies I build custom software for companies across the US, Canada, the Gulf, Europe and Australia.
Player profile
I'm a hands-on solution architect and full-stack AI engineer with 8+ years of designing, building and shipping production software. My work runs from Python (Django, Flask) and JavaScript (React, Node.js) through to retrieval-augmented generation, LLM applications and multi-agent systems.
Recent builds include a graph-based RAG legal assistant for Dutch law firms that raised EUR 1.2M in pre-seed funding, a bilingual Arabic and English legal AI platform hosted inside the client's own Azure tenant, and a LangChain and Memgraph multi-agent financial research platform with automated LLM evaluation.
Across Aeyron and my agency work I've helped 140+ companies build their products, deployed 300+ web and mobile apps, and shipped AI products that reach more than 7 million users. I've also published research on hosting LLMs on-premise with highly quantized models. I'm equally comfortable leading a team or working as a dedicated engineer inside a distributed client team.
Inventory
Quest log
NeuraVue Ltd.
Stealth Startup
MEShBI
BikiaHealth
Sesame
Dubai Silicon Oasis
EARLIER SAVES: Computer Vision Engineer, Stagnoli (Italy, 2020) · AI & IoT Intern, Software Productivity Strategists (2019) · Intern, DynamoLogic Solutions (2018)
Level select
7 levels
Bilingual Arabic and English RAG platform for Saudi-law research, drafting and due diligence, hosted in Azure with document-repository integrations and governance controls.
Graph-based RAG legal assistant for Dutch law firms that learns from lawyer feedback.
Multi-agent financial research over SEC filings and market data, with automated hallucination and quality evaluation.
Extracts patient data from faxes, reports and scans into structured records with confidence scoring and human verification.
On-premise meeting assistant for Teams, Google Meet and Zoom with English, Arabic and Spanish transcription, summaries and chapters.
AI voice agent that calls suppliers, collects quotes and compares them in one dashboard.
AI medical education with OSCE clinical simulation, adaptive assessments and knowledge search.
6 levels
Telehealth, patient records, in-person care and hospital logistics in one secure system with operational analytics.
Loyalty and payments with QR, NFC and cash, plus customer app, merchant web app and admin panel.
Procurement, vendor, recipe and packaging management for a large food kiosk in Islamabad.
AI trading-card grading app with marketplace, social features, real-time chat and a revenue dashboard.
Maritime platform for vessel tracking, inventory, inspections, HSE alerts and compliance.
Online fundus examination with AI results, doctor review and downloadable patient reports.
Bonus stages
Video analytics for bay utilisation, service duration and exceptions, with AI-assisted querying.
Multi-site wait-time, dwell-time and zone analytics running on Nvidia Jetson.
Side quest
IEEE · PUBLISHED PAPER
Research on running large language models inside an organisation's own infrastructure by compressing them to very low bit widths, so private data never has to leave the building. The same thinking drives MedAide's data-sovereign deployments and the on-premise meeting assistant.
READ ON IEEE XPLOREHall of fame
| AWARD | FROM | YEAR |
|---|---|---|
| Gold Medal, Computer Engineering, Class of 2021 | NUST | 2021 |
| National Winner (Pakistan), Imagine Cup | Microsoft | 2021 |
| 50 Under 50 Entrepreneurs | NUST | — |
| Top Rated Plus · Top 1% Islamabad, Top 3% globally | Upwork | NOW |
| Data Science using Python | IBM CognitiveClass | — |
| TensorFlow for Deep Learning | Certification | — |
Continue?
Building a RAG assistant, an AI agent, or a computer vision system that has to run on real hardware? Send a short note on what you're working on and I'll reply with next steps.