< MLOps Engineer />

Muhammad Dawood Khan

Building ML systems that ship to production — not just notebooks.

python train.py --pipeline production
Scroll to explore

About

Muhammad Dawood Khan

Muhammad Dawood Khan is an MLOps and Machine Learning Engineer building and deploying production-grade ML systems. From experiment tracking with MLflow and DVC to containerized deployments on Railway, his work spans the full ML lifecycle.

He has shipped a live property valuation API, contributed ML models to PhD-level AI labor economics research, and built medical imaging pipelines achieving 0.92 Dice score. He specializes in making models production-ready — not just accurate.

0.00
Dice Score (Medical Imaging)
Live on Railway
Property Valuation API
0
Production Projects Shipped
Active Collaborator
PhD AI Labor Market Thesis
ARSENAL

Core Competencies

MLOps & Deployment

MLflow DVC Docker FastAPI Railway CI/CD GitHub Actions uv

ML/DL Frameworks

PyTorch TensorFlow Scikit-learn LightGBM XGBoost CatBoost Optuna TabNet

Languages & Infra

Python Bash PostgreSQL Ubuntu Linux Docker Compose

Data Engineering

Pandas NumPy Feature Engineering Quantile Regression Bayesian Target Encoding

Shipped Work

Live on Railway

propval-pk

Pakistan Property Valuation API. End-to-end MLOps pipeline for real estate price prediction. Integrates Bayesian target encoding and quantile regression for uncertainty-aware predictions.

LightGBM MLflow DVC FastAPI Docker Railway
0.92 Dice Score

Brain Tumor Segmentation

Medical Imaging pipeline. Implements 2D MRI segmentation using a customized ResNet-34 semantic segmentation model with specialized data augmentations and mixed-precision training.

PyTorch ResNet-34 Segmentation
Deployed

textile-scheduler

Production Web App. A full-stack monorepo, end-to-end deployed, solving CORS, containerized caching and builds, and establishing a robust production release pipeline.

FastAPI React Vite Docker Railway
Competition Pipeline

SOFTEC 2025 ML Competition

Ensemble Pipeline for predictive ML. Engineered a six-model ensemble framework with automated Optuna hyperparameter sweeps and integrated TabNet/CatBoost estimators.

LightGBM XGBoost CatBoost TabNet PyTorch Optuna
Active

PhD Thesis Collaboration

AI Labor Economics. Led machine learning and statistical modeling (Random Forest & Logistic Regression) to analyze the impact of AI adoption on twin cities' IT labor market.

Scikit-learn Statsmodels Random Forest Logit Regression
Client Project

Image Classification Pipeline

Delivered a highly reproducible image classification pipeline fine-tuning ResNet models on STL-10, with automated evaluation reporting and structured hydra configuration management.

PyTorch ResNet Transfer Learning

Experience

PhD Thesis Collaboration (ML & Econometrics)

2025 – Present

Executing machine learning and econometric modeling (Random Forest, Logistic Regression, K-Means Clustering) on survey data from IT firms to analyze labor market dynamics, reskilling needs, and job displacement risks.

Freelance Machine Learning Engineer

2024 – Present

Contracted by diverse clients to design and deploy specialized ML models. Delivered scalable image classification tools, engineered reproducible pipelines using configuration frameworks, and implemented automated performance logging systems.

Let's Build Something

Open to MLOps, ML Engineering, and Data Science opportunities. Based in Pakistan, available remotely. Let's design something scalable.