Hi, I'm Zulfiqar

I am an AI Engineer and Data Scientist driven by a passion for multimodal intelligence, deep learning, and statistical modeling. My work focuses on transforming complex data into meaningful insights and practical solutions across diverse domains, while my research delves into Generative AI, Trustworthy AI, and Human-AI collaboration, aiming to advance methods that connect innovation with real-world impact.

See My Resume

What I Do

Innovative Solutions for Real-World Problems

Data Science & AI

🧠 Generative AI and Large Language models(LLMs)

🩺 Computational Neuroscience

🔍 Time-series forecasting & Bayesian modeling

📊 Real-time analytics & monitoring

Full-Stack & Cloud

🚀 SaaS Applications with FastAPI & Next.js

🌐 RESTful APIs & real-time dashboards

💾 NoSQL / SQL / Vector Databases

Proficiency

Deep Learning
95%
Generative AI
93%
Research & Development
90%
Python
95%
PyTorch & HF
92%

Education

COMSATS University Islamabad
M.S. Mathematics (Machine Learning Focus)
Sept 2018 – Sept 2021

Thesis: “Child Behavior Classification using Machine Learning” – predictive modeling on health time-series data.

    University of Gujrat
    B.S. Mathematics
    Sept 2012 – June 2017

    course work in applied mathematics & statistics and computer science.

      Experience

      /img/icons/common/demp.png

      Ministry of Information & Broadcasting – DEMP

      AI Engineer
      August 2025 – Present
      Real-time TTS & ASR for broadcast media; transformer pipelines; AI-driven multimedia analytics dashboards.
        /img/icons/common/pid.jpg

        Press Information Department, MOIB

        Data Analyst / Applied Researcher
        March 2023 – July 2025
        Fine-tuned Llama-2-7B (4-bit QLoRA) on 4 M social posts (hate-speech F1=0.89); built live social-media streams & policy dashboards.
          /img/icons/common/Sejong_University.png

          Sejong University – Optical Eng.

          Research Assistant (Remote)
          Sept 2022 – March 2023
          U-Net OCT segmentation; hyperbolic & graph neural nets for cardiovascular diagnostics.
            /img/icons/common/ncai.png

            NCAI SmartCity Lab

            Deep Learning Engineer
            Aug 2021 – July 2022
            Computer-vision safety stack for AVs: driver-behavior monitoring & object detection with policy-driven urban mobility.
              /img/icons/common/comsats.png

              COMSATS University

              Graduate Research Assistant
              Sept 2017 – Aug 2021
              Hierarchical Bayesian models for survey data; semi-supervised EM imputation; behavioral research frameworks.

                Projects

                AI Support Assistant

                AI Support Assistant is a lightweight, floating chatbot widget built with Next.js, React, and TailwindCSS, powered by Gemini API + Qdrant vector search. It lets users upload documents, ingest URLs, and chat with an AI assistant in real time, providing contextual support inside any web app.

                Demo

                Multimodal Emotion Recognition

                Attention-based fusion of facial & speech signals using CNN+LSTM (PyTorch) for mental-health analytics.

                Demo

                NextGen-ai Generative SaaS

                Multimodal SaaS (text, image, audio) via GPT-4-V + Stable Diffusion; 150 k API calls/mo; LoRA fine-tuning (8× speed-up).

                Demo

                AI-News Summarizer & Sentiment Engine

                Real-time GPT-based news summarization & sentiment extraction for MOIB policy dashboards.

                Demo

                3D Biomedical Image Segmentation

                U-Net & 3D CNNs for semantic segmentation of tumor regions in volumetric scans to boost diagnostic accuracy.

                Demo

                Audio-Visual Speech Recognition

                Whisper + lip-reading module for robust ASR in noisy environments; CUDA-accelerated multi-GPU batching.

                Demo

                Child Behavior Classification

                ML framework to classify child behavior patterns from longitudinal health data (time-series). MSc thesis project.

                Demo

                Our Clients Feedback

                Dr. Sarah Kim

                Professor of Biomedical Engineering

                Zulfiqar’s innovations in biomedical imaging have significantly advanced our research capabilities.

                Mohammed Khan

                CTO, InnvoAI

                Zulfiqar delivers complex AI solutions under tight deadlines—an invaluable team asset.

                Get in Touch!

                I am currently seeking fresh opportunities and welcome messages in my inbox. Whether you have inquiries, job prospects, or simply want to connect, I ' ll make sure to respond promptly!

                Data Scientist||Researcher

                Islamabad Pakistan