Richhiey Thomas

Hey there! Richhiey here.

I design and deliver end-to-end machine learning systems, specializing in audio ML — automatic music transcription, event detection, and ASR. My approach spans building robust data pipelines, optimizing real-time audio models, and deploying production-grade solutions on the cloud. I leverage Python, TypeScript, and C++, alongside ML frameworks and audio software to transform complex data into actionable results.

My cloud skillset includes AWS (SageMaker, Glue, S3, ECS), Docker-based containerization, and rapid API development with FastAPI and Flask. For scalable, reliable data workflows, I use PySpark for distributed processing and Airflow for orchestration, ensuring maintainable and efficient ML pipelines that consistently meet business requirements.

If you’re interested in ML for audio, robust data engineering, or simply want to discuss how to build production AI from the ground up — let’s connect. This site showcases a sample of my projects, thoughts, and learnings. Thanks for visiting!