CodeGuard AITeam lead
A code-review application that brings together static analysis, GitHub integration and structured developer feedback. I led the team and coordinated development, integration, testing and presentations.
Software engineer in Karachi building Python APIs, applied AI and automation workflows with FastAPI, Docker and AWS.
I enjoy the work that connects an idea to a functioning application: building the API, integrating services, handling data and debugging the parts that do not work as expected.
At Traders Global Generation, I am developing a social platform with an integrated marketplace: Next.js frontends for admins, moderators and users, an Expo and React Native mobile app in TestFlight testing, Redis-based trend tracking and AWS hosting. I also support software quality assurance on an outsourced project. Previously, I worked on a Python and FFmpeg sports-highlight pipeline at Tapmad Entertainment and built API-connected Python tools at Websity Digital.
My interest in applied AI has led me to team projects in scholarship discovery and code review, as well as an ongoing final-year project combining medical-image analysis with recommendations grounded in curated information. I was recognised as a Top Performer in the Pak Angels / ASPIRE Generative and Agentic AI training programme, and have completed six Claude Academy courses and the n8n Foundations pathway.
I am completing my BS Computer Science at IoBM, with graduation expected in January 2027. I want to keep developing my skills in backend engineering, cloud-native applications and reliable AI systems.
Grouped the way I actually reach for it, not by how long the list looks.
Four here: two team builds I led, a media pipeline and my final-year system. The other twelve are listed below.
A code-review application that brings together static analysis, GitHub integration and structured developer feedback. I led the team and coordinated development, integration, testing and presentations.
A scholarship discovery application combining search, eligibility checks and document-processing workflows, with a FastAPI backend behind a Next.js interface.
A Python and FFmpeg workflow for turning video streams into clips and sports-highlight candidates. A FastAPI backend coordinates processing; I built and integrated the processing components and automated the workflow steps.
My final-year project: a mobile system pairing knee X-ray analysis with personalised recommendations grounded in curated physiotherapy information. An academic prototype under development.
Twelve more, from a Hadoop pipeline to a hand-sign classifier, in the archive below.
Current work at Traders Global Generation: a social platform with marketplace features. Next.js frontends for admins, moderators and users, an Expo / React Native mobile app in TestFlight testing, Redis-based trend tracking and AWS hosting.
A Django application I built, containerised and deployed to EC2, handling the deployment configuration, networking and the issues along the way.
A production server on AWS EC2 running custom-compiled, hardware-accelerated video runtimes under Docker Compose, behind an Nginx reverse proxy, writing to locked-down object storage.
Batch processing over large e-commerce datasets in Hadoop and Hive. Reworking the job structure and partitioning cut total execution time by about 40%.
An AI-assisted platform that ingests QuickBooks and Faire data, maps it automatically, and runs the ETL and reporting asynchronously through Celery and Redis.
Real-time hand-sign recognition on a custom landmark dataset I collected myself, holding a steady 30 FPS. Trained in Colab, runs on a webcam.
An ad-insertion panel for live production, driven by a FastAPI backend over local network protocols with full runtime activity logging.
Scrapes, cleans and charts client data in one Streamlit app, which turned proposal writing for 100+ clients into a two-minute job.
Pulls repository metrics through the GitHub API, runs them past Gemini for a written read, and files the results automatically.
A DALL·E 3 front end for the content team, built to keep image output flowing without hitting a shared ChatGPT Pro ceiling.
A ChatGPT-style client with streaming responses and persisted threads, written in TypeScript against the OpenAI API.
Takes raw phone-number records and runs them through an automated enrichment and validation pipeline before handoff.
A junior role, a technical administrator role and two internships, most recent first.
Two training programmes, a learning pathway, ten course certificates and three event certificates. Entries marked Verify link to pages where you can check them yourself.
Issued by n8n GmbH for completing all three Foundations courses (n8n101 – n8n103): building, integrating, testing and running production workflows, including AI agents.
n8n Academy course completion certificate.
n8n Academy course completion certificate.
n8n Academy course completion certificate.
n8n Academy course completion certificate.
Claude Academy course completion badge.
Issued by ASPIRE Pakistan, Cohort 11 (Aug – Oct 2026). Generative AI app development, agentic and multi-agent workflows, RAG, vector retrieval and prompt engineering. I led the team on ScholarCompass and CodeGuard AI.
Claude Academy course completion badge.
Claude Academy course completion badge.
Claude Academy course completion badge.
Claude Academy course completion badge.
Claude Academy course completion badge.
Certificate of participation in the final-year-project defense module at TechNova '26, IoBM, Karachi.
Training and internship completion, held by Plus W Inc. and OEC through the Pakistan Japan Centre for Mutual Cooperation (Jan – Dec 2025).
Certificate of participation from Backdoor Security Agency, for the cyber security seminar held at NASTP, Karachi.
Certificate of attendance for the Learning Teaching Training Activities organised by ODTÜ, MEBIVA, EKAPDER and the Ministry of Agriculture & Forestry, Ankara (Oct 11 – 15, 2024).
Results I am happy to be asked about in an interview.