Multi-Agent E-Commerce AI Platform
Autonomous Customer Support, Sales and Purchase agents with real-time marketplace integrations (BackMarket, Refurbed, C-Discount) and a Next.js chat frontend.
Available for freelance & full-time AI engineering roles
Hi, I'm Muhammad Haseeb Aslam
Senior Full Stack AI Developer — |
5+ years building production-grade AI systems — from multi-agent LLM platforms and RAG chatbots to computer-vision diagnostics and Next.js front ends that put them in front of users. Published researcher, ex-Lead of AI Products at a healthcare-AI startup.
0+
Years Experience
0+
AI Systems Shipped
0
IEEE Publications
0%
Fraud Model Accuracy
About Me
Turning AI research into shipped products
I'm a Senior Full Stack AI Developer and Data Scientist based in Gojra, Pakistan, with 5+ years of experience across Machine Learning, Deep Learning, Generative AI, and Large Language Models. I specialize in taking AI from research notebooks to production — healthcare fraud detection, medical diagnostics, NLP document intelligence, and e-commerce automation.
On the backend I build multi-agent LLM systems, RAG pipelines, and scalable ML services with Python, FastAPI, Docker and Kubernetes across AWS, Azure and GCP. On the frontend I ship the interfaces those systems live behind — real-time, streaming chat UIs and dashboards built with Next.js and React, so the AI is actually usable, not just accurate.
Previously I led AI product development at a healthcare-AI startup and conducted deep learning research at the GIK Institute's Machine Learning & Data Science Lab, publishing 2 papers at IEEE ICET 2022.
AI
LLMs
RAG
Next.js
CV
MLOps
Toolbox
Skills & Technologies
Full stack across the AI product lifecycle — from model research to the interface users touch.
Generative AI & LLMs
Frontend Development
Backend & APIs
ML / DL Frameworks
Computer Vision & NLP
MLOps & Cloud
Career Path
Professional Experience
Senior AI Engineer & Consultant
Aug 2025 – PresentE-Commerce AI Solutions (SmartTecAnalytics) · Remote
- Architected a multi-agent AI system (Customer Support, Sales, Purchase agents) for a large-scale e-commerce platform, cutting manual workload by 60%.
- Built a RAG-based Customer Support Agent resolving 80% of queries autonomously, slashing average response time from 12 hours to under 2 minutes.
- Engineered LangChain + GPT-based chatbots per agent, reaching 90%+ customer satisfaction scores.
- Integrated real-time data pipelines from 5+ international marketplaces (Back Market, Refurbed, C-Discount and more) for unified catalog management and pricing intelligence.
- Built a Sales Agent with marketplace analytics and recommendation logic, lifting conversion rates by 25%.
- Shipped the customer-facing chat interfaces in Next.js and React with real-time streaming, multi-language support and seamless agent handoff, backed by Docker/FastAPI CI/CD pipelines.
Senior Data Scientist — Lead AI Products
Sep 2023 – Aug 2025SeethruTec · Lahore, Pakistan
- Led development of 8+ AI-driven healthcare products, owning ML pipelines from data collection to production deployment.
- Built a deep-learning OCR system for medical records, cutting manual processing time by 70%.
- Developed an LLM-based AI-generated text detector for medical records at 95% accuracy, ensuring compliance.
- Designed and deployed Alivia GPT, a virtual assistant platform for building custom, file-grounded AI assistants using RAG.
- Implemented AI-driven X-ray duplicate detection, reducing redundant records by 45%.
- Built a CNN-based cardiac stenosis detection system achieving 93% sensitivity on heart X-rays, plus an immersive fraud-analytics visualization for Apple Vision Pro.
Research Assistant
Feb 2021 – Jun 2023Machine Learning & Data Science (MDS) Lab, GIK Institute · Topi, Pakistan
- Published 2 papers at IEEE ICET 2022 on deep learning for medical image classification.
- Built a fatty liver disease severity model reaching 99% accuracy using SVM, Random Forest and ensembles.
- Designed 3D-CNN / C_UNet architectures for hyperspectral image segmentation.
- Built malware detection using mutual-information feature selection, achieving 98% accuracy with optimized KNN.
- Built intrusion detection systems using ML and deep learning to analyze network traffic and detect cybersecurity threats.
Selected Work
Featured Projects
Client and research work — most under NDA, shown here without public repos.
Medical Image Classification
Multi-disease diagnosis system: 96% pneumonia detection, 94% brain tumor classification, 91% skin cancer detection, with Grad-CAM explainability.
Real-Time Fraud Detection
Streaming pipeline processing 10,000+ transactions/minute at 99.2% accuracy and <0.1% false-positive rate using ensemble ML.
NLP Healthcare Document Analysis
Medical NER at 95% F1, automated ICD-10 coding at 92% accuracy, and transformer-based patient outcome prediction.
Intelligent Passenger Estimation
Real-time vehicle occupancy detection at 92% accuracy for traffic management, deployed via Docker + FastAPI.
Industrial IoT Anomaly Detection
Unsupervised predictive maintenance for industrial equipment at 94% precision, deployed to AWS IoT edge devices.
Manufacturing Quality Control System
Automated defect-detection system for production lines, hitting 97% accuracy while processing 500+ items/minute via industrial cameras.
Research
Publications
IEEE ICET 2022
↗
IEEE ICET 2022
↗
Prediction of Heart Failure by using Machine Learning and Feature Selection
M.H. Aslam, S.F. Hussain · 17th International Conference on Emerging Technologies (ICET), IEEE
Predictive Analysis on Severity of Non-Alcoholic Fatty Liver Disease (NAFLD) using Machine Learning Algorithms
M.H. Aslam, S.F. Hussain, R.H. Ali · 17th ICET, IEEE
🎓 Education
MS, Computer Science
GIK Institute of Engineering Sciences & Technology · 2021–2023
Thesis: Classification & Semantic Segmentation of Hyperspectral Images via 3D-CNN & C_UNet
BS, Computer Science
Bahria University, Lahore · 2016–2020
FYP: Detecting Lung Cancer using Image Processing
🏆 Leadership
President, Graduate Student Society
GIK Institute
Organized international conferences and managed Pakistan's largest ICPC programming competition, leading a team of 30+ students.
Get In Touch
Let's build something intelligent
Open to freelance engagements, full-time roles, and interesting AI problems. Reach out.