Qalab Hassnain AghaCo-Founder & CTO
ResumeBook a call

About

Qalab Hassnain Agha

Co-Founder & CTO · AI Systems Architect · Islamabad, Pakistan

I'm the Co-Founder & CTO of Quickgen Technologies, where I architect and personally code the systems we ship, from first prototype to production, and lead a team of 25+ engineers. Over 6+ years I've built AI, IoT and full-stack products for clients in the UAE, UK, Australia and Pakistan, including QuickComm, the voice AI platform that replaced hotel walkie-talkies and won 5 P@SHA ICT Awards in 2026. I care less about which model is in fashion and more about the latency budgets, cost ceilings and monitoring that keep AI working after launch.

Find me on

Where I work

Co-Founder & CTO, Quickgen Technologies

An AI and IoT engineering company incubated at the National Science and Technology Park (NSTP), NUST, Islamabad. Quickgen builds its own products, including QuickComm AI and Scentix, and engineers AI systems for clients in Australia, the UAE and the UK.

You can hire me directly for consulting, architecture and fractional CTO work, or engage Quickgen when you need a full team behind the build.

Education
MS Computer Science, Sir Syed CASE Institute of Technology, Islamabad 2023-2026
B.E. Computer Engineering, National University of Sciences and Technology (NUST) 2017-2021
Publications
Edge AI-Enabled Intelligent Healthcare Framework Using Lightweight Deep Learning, Multimodal Sensor Fusion, and Real-Time On-Device Analytics for Continuous Patient Monitoring and Early Health Risk Detection International Journal of Advanced Research (IJAR) (HEC-recognised journal, Y category), November 2026. First and corresponding author.
“Wait, Cancel That”: Release Monotonicity and Cancellation Semantics for Streaming Voice-Agent Actions Under review at Computer Speech & Language (2026). Sole author.
Feature Selection and Model Optimization for Accurate Classification of Complex Data Journal of Global Social Transformation (JGST), October 2026. First author.
Beyond Automation: Generative AI’s Influence on Human Decision-Making, Trust, and Accountability in Intelligent Cybersecurity Systems International Journal of Advanced Research (IJAR) (HEC-recognised journal, Y category), October 2026. Co-author. My takeaways
What Transfers From Synthetic to Real Scalp Dermoscopy? A Pre-Deployment Component-Level Evaluation for Handheld Trichoscopy Submitted to Computer Methods and Programs in Biomedicine (2026). Preprint. Sole author. Dataset: ScalpHairLabels v1 (504 annotated scalp dermoscopy tiles, CC BY 4.0). Code: Analysis code (reproduces every reported result, MIT).
Certifications
Neural Networks and Deep Learning Coursera
AI for Medical Diagnosis Coursera
Applied Data Science with Python Specialization Coursera

What I build with

  • Voice AIStreaming speech pipelines that hold up on noisy, real-world audio.Whisper · Deepgram · ElevenLabs · PCM audio streaming94%+ transcription accuracy, <800ms voice round trip
  • Computer VisionDetection, tracking and alerting on live camera feeds.YOLOv8 · OpenCV · TensorFlow91% accuracy across 8+ live feeds
  • AI & LLM SystemsLLM features, RAG and agents with guardrails, evals and cost ceilings.Gemini · OpenAI · Claude · RAG & vector search~88% intent precision in production
  • Real-Time & IoTSensor and device data streamed to dashboards in milliseconds.WebSockets · MQTT · BLE 5.0 · ESP32200Hz+ ingestion, sub-100ms latency, 500+ sessions
  • Backend, Cloud & ProductThe services, data and apps that carry the AI to real users.Python · FastAPI · PostgreSQL · Redis · AWS · Azure · Docker · Next.js3× throughput after a microservices migration
How I apply these for clients

Independent recognition

6 P@SHA ICT Awards

QuickComm AI, the system I architected and coded, took five of them. Judged by an industry jury of the Pakistan Software Houses Association, the national body for Pakistan’s IT industry.

The full story from awards night

04Experience

I design the architecture. I write the code.

Every system in this portfolio is architected by me and built with my own hands on the critical code. Here is what that ownership looks like in numbers.

  • 3×throughputArchitecture decisionsI draw service boundaries around the real load profile. I designed and led QuickComm’s move from a monolith to AWS microservices: 3× throughput with near-zero deployment downtime.
  • ~$3/moinfra per propertyCost and performance ownershipBudgets are set before code: roughly $3/month of infrastructure per hotel property, under 800ms for voice, under 100ms from sensor to dashboard.
  • −70%incident response timeProduction responsibilityAutomated anomaly detection on service health metrics cut critical incident response time by 70%.
  • Every weekI write and review production codeHands on the codeI architect the systems and write the critical paths myself: the QuickComm voice pipeline, the uGlowMD delivery layer and the Wild Hive stack are my code. I also lead 25+ engineers at Quickgen.

Co-Founder & Chief Technology Officer

Quickgen Technologies

2022 to PresentIslamabad, Pakistan · clients in Australia, UAE and UK

  • I own the architecture and write the critical code: technical strategy and end-to-end delivery across AI, IoT and SaaS products in healthcare, hospitality, fintech and consumer tech.
  • Built a YOLOv8 CCTV anomaly detection system at 91% accuracy across 8+ simultaneous camera feeds, cutting false alerts by 35%.
  • Built the Apollo Golf GPS backend serving 1,000+ active users with <80ms queries and ±2m GPS accuracy.
  • Lead a team of 25+ engineers across the full lifecycle, from system design through deployment and post-launch iteration.

Products and client engagements

QuickComm AECo-Founder & CTO

Quickgen product · Dubai, UAE

5 P@SHA ICT Awards, 2026
  • Co-founded QuickComm, then architected and coded it myself from first prototype to production.
  • Real-time voice pipeline replacing hotel walkie-talkies: 94%+ transcription accuracy, ~88% intent precision, ~45% faster staff response.
  • Led the monolith-to-microservices migration on AWS: 3× throughput, near-zero deployment downtime.
  • Cut critical incident response time by 70% with automated anomaly detection on service health metrics.

upLYFTIoT & full-stack lead · contract

Client engagement · London, UK (remote) · 2024 to present

  • Designed the ingestion and analytics architecture and built the backend and web app for a two-sided physiotherapy platform, cutting clinician onboarding time by ~60%.
  • Built 200Hz+ sensor ingestion over WebSockets and UDP on Azure: sub-100ms latency, 500+ concurrent sessions, 92%+ movement classification accuracy.
  • Shipped 3 major releases on schedule as technical lead of a 5-person team.

Before Quickgen

  • AI Engineer & Backend Developer, Freelance (Upwork & direct)2021 to 2022Delivered 15+ AI and backend projects for international clients: OCR, pose estimation, object detection, NLP automation and full-stack web apps.
  • Information Technology Intern, CareCloudMay 2021 to Aug 2021Built health-services REST APIs in .NET Core for a live healthcare production system.
  • Software Engineer Intern, PTCLAug 2019 to Sep 2019Built an employee record search desktop app for HR using Python and deep learning.
Full resume (PDF)

Have an AI system that needs to survive production?

Tell me what you’re building. In 30 minutes you’ll get an honest read on the architecture, the risks and what I would build first. No commitment, no sales pitch.

Book a 30-min call Send a messagesupport@qalabagha.com