ML pipelines that actually ship.

Machine learning engineer building production-grade MLOps systems end to end.

machine learning mlops data engineering
Highlights

JobSwipe — AI-Powered Job Search Assistant

Multi-agent AI platform with a Tinder-style swipe UI: upload your résumé, get role recommendations, identify skill gaps, and auto-apply — all orchestrated by four LangGraph agents with FAISS semantic job matching.

Live ↗ Python LangGraph Next.js FAISS Flask Supabase

JobSwipe — Job Board Scraper & Dashboard

Daily scraper pulling 1,000+ listings across LinkedIn, Indeed, Rozee & Mustakbil, with Redis deduplication, GitHub Actions scheduling, and a Next.js dashboard for browsing and filtering all jobs in real time.

Live ↗ Python Scrapling Next.js Supabase Redis GitHub Actions
ML & MLOps

Telecom Churn MLOps Pipeline

Seven-model benchmark (XGBoost, LightGBM, Stacking Ensemble, and more) with MLflow experiment tracking, FastAPI serving, Prometheus metrics, Grafana dashboards, and a drift simulation workflow — all containerised with Docker Compose.

Python MLflow FastAPI Docker XGBoost Prometheus

Breast Cancer Detection — Kubeflow ML Pipeline

End-to-end classification pipeline on Kubeflow Pipelines running on a local Kubernetes (Minikube) cluster, featuring Genetic Algorithm–tuned SVM (98.25% accuracy, 99.60% AUC-ROC) and Particle Swarm Optimization–tuned Neural Networks across four parameterised pipeline versions.

Python Kubeflow Kubernetes scikit-learn DEAP

FMA Music Recommendation Engine

KMeans clustering on 15 audio features (MFCC, tempo, spectral centroid) via Databricks medallion architecture (bronze→silver→gold), with real-time Kafka streaming, MLflow model registry, and a Flask UI backed by AWS S3 and the Databricks SQL Connector.

Python Databricks Apache Spark Kafka MLflow Flask

Automated Drift-Aware Fare Predictor

Production MLOps pipeline for NYC taxi fare prediction with automated monthly data ingestion via Airflow, PSI-based statistical drift detection, Champion vs Challenger model promotion, and a LightGBM model served as a REST API — fully containerised.

Python Airflow MLflow Docker LightGBM PostgreSQL

Short-Term Load Forecasting — End-to-End MLOps

Energy demand forecasting with XGBoost trained on 1.3M rows of EIA930 data, deployed via FastAPI on AWS App Runner with a CI/CD pipeline through GitHub Actions and Docker — from raw multi-source ETL to a live prediction dashboard.

Python XGBoost FastAPI Docker AWS App Runner GitHub Actions

I'm a Machine Learning Engineer focused on systems thinking and building production-grade systems. It's easy to obsess over a model's accuracy, but the real work lies in targeting actual business metrics and surviving data drift in production. I specialize in the serving side of AI, building robust ETLs and deploying scalable infrastructure using AWS and core MLOps tools. For me, a project is only truly done when it's actively serving users and handling real-world load without breaking. I am a fresh FAST Islamabad graduate with Cum Laude honors, driven by the challenge of turning complex data into reliable systems. Outside the terminal, you'll find me watching football, jogging, or occasionally swimming.

I've got hands-on experience with…

Python (Pandas, NumPy, Scikit-learn, OpenCV, Pytorch, Tensorflow) AWS Apache Airflow Apache Kafka Apache Spark Apache Hadoop MLflow Kubeflow Docker FastAPI PostgreSQL Databricks Snowflake CI/CD Pipeline Tableau Prep and Tableau Desktop d3.js C/C++
May 2026 – June 2026 AI Software Engineer Intern @ Dafinitiq AI
Python · Large Language Models · Voice AI Agent · STT · TTS
Mar 2026 – June 2026 Database Development Intern @ Inzent
Oracle · PostgreSQL · C · OCL · ODBC
Aug 2023 – May 2026 Teaching Assistant @ FAST Islamabad
Parallel and Distributed Computing, Operating Systems, Data Analysis, Fundamentals of Software Engineering, Discrete Structures · 100+ students/semester
Feb 2025 – May 2025 AI Engineer (Contract) @ DMLYTICS
Orthomosaic pipeline · OpenCV · GDAL · Precision agriculture