AhsanQazi

Senior Full Stack AI Engineer

Zavtech AI

Aug 2025 - Present • 1 yr 1 mo

Built and shipped SaaS products end to end, using Next.js frontends with Django backends—from system architecture through production deployment. Developed a UK-focused online marketplace using Django and Next.js, featuring Stripe Connect escrow payments, real-time messaging through Django Channels and Redis, Celery-based payment release scheduling, and subscription billing. Implemented AI-powered identity verification and trust scoring, including email, phone, and document verification. Built an identity verification product featuring face recognition and matching, document OCR, and AML screening, delivering it from architecture through deployment. Developed an HR SaaS product with an Angular frontend and Django REST backend.

Software Engineer

IIFA TECH

Jun 2024 - Jul 2025 • 1 yr 1 mo

Data science and full-stack delivery across client projects. Built Pricify, a pricing analytics platform for the UK market — Django and React, where customers upload their own datasets and receive analysis back as interactive charts. Modelled price elasticity of demand using Bayesian analysis, translating statistical output into visualisations non-technical users could act on directly. Developed a heart disease risk prediction system using machine learning and deep learning models on medical datasets, assessing clinical factors including age, cholesterol and ECG readings. Built with Scikit-learn and TensorFlow, with a React frontend for visualising risk factors. Built backend services in FastAPI for a property platform.

Associate Data Scientist

Programmers Force

May 2023 - May 2024 • 1 yr

Data science work on Shufti, a London-headquartered KYC/AML identity verification platform serving regulated businesses globally. Trained and evaluated computer vision models for identity document OCR, extracting structured fields from ID documents across varied formats and languages. Worked on face recognition and face matching, verifying selfie captures against document photographs. Contributed to deepfake and presentation-attack detection to flag spoofed verification attempts. Supported AML screening, including name matching and risk flagging against sanctions and watchlist data. Built data preparation and model training pipelines in Python (Pandas, NumPy, Scikit-learn) with PyTorch and TensorFlow.