Mohammed Niyas

Hey there! I'm Niyas, an ML engineer with a passion for creating data and machine learning products that make a difference. My interests span from art and engineering to finance and history, and I love weaving these into my projects. I'm glad you’re here—explore around, and I hope you find something that sparks your interest!

Talks, Thoughts & Works

Mapping India
A State-by-State Look at Coronavirus in US Prisons
Becoming a good manager

Skills

Every day, I dive into creating data pipelines, building ML services, and developing tools that make machine learning more accessible and impactful. I’m passionate about taking complex processes—like data acquisition, ETL/ELT, model training, and deployment—and turning them into streamlined, efficient pipelines. My goal is to ensure everything runs smoothly, from monitoring models to deploying dashboards, with a focus on automation that saves time and enhances results.
ML/DL
I have worked on building and deploying ML models for over 4 years. I have experience in building models using libraries like PyTorch, HF-transformers, Scikit-learn, and FastAI. I've also deployed models using Triton Inference Server, TorchServe, BentoML and FastAPI. I have experience in building models for NLP, Computer Vision, and Tabular data.
DevOps/MLOps
I’ve worked on deploying ML services and web apps, setting up databases, and building data pipelines across diverse environments. Along the way, I’ve created CI/CD and MLOps pipelines using GitLab, Bitbucket, and GitHub. I’ve also set up orchestration tools like Airflow and Dagster. In my day-to-day work, I rely on tools like Docker, Kubernetes, and Terraform to keep everything running smoothly.
Python
My "native" programming language, I've worked with it for over 9 years. I've used it for building web apps, performing data analysis, scraping websites, developing ETL pipelines, building Machine Learning and Deep Learning models, creating application tests suites and building CI/CD pipelines. I use FastAPI, PyTorch, Pandas, Numpy, Pytest libraries on my day to day work.
Database
I’ve been working with relational databases for over 8 years, often alongside Python, to build data-driven applications. For general-purpose use cases, I primarily turn to PostgreSQL, while I’ve utilized Redshift and BigQuery for data warehousing needs. I’ve also had experience with Redis, MongoDB, and ElasticSearch on various projects.