Skip to content
View S-Houston's full-sized avatar

Block or report S-Houston

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
S-Houston/README.md

LinkedIn GitHub followers

👋 Hi, I’m Stuart Houston

Senior-level Data Scientist | MLOps & AI Infrastructure

I’m a Data Scientist focused on building production-ready, modular ML pipelines that demonstrate value quickly while laying the foundations for scalable, secure deployment.

As the sole Data Scientist at CNOOC International, my work involves more than just model development it’s about proving what’s possible, showing where Data Science creates value, and designing workflows that can evolve smoothly from proof-of-concept to production once the infrastructure is ready.

Much of my work uses controlled-access industrial data governed by internal standards, so the public repositories here serve as case studies that reflect similar challenges: messy, real-world data, feasibility testing, and designing systems that are robust enough to scale.


🔧 Featured Projects

  • NASA CMAPSS dataset → Remaining Useful Life prediction
  • End‑to‑end workflow: feature engineering → model training (LightGBM/XGBoost) → MLflow tracking → FastAPI service → Streamlit dashboard
  • Structured with Cookiecutter DS template and clear modular organisation to support testing and reproducible workflows.
  • Business framing: risk classification (high/medium/low) for asset management decisions
  • Computer vision + OCR pipeline for legacy engineering diagrams
  • YOLOv5 for symbol detection, EAST for text detection, Tesseract for OCR
  • Streamlit app for interactive validation
  • Evaluation with precision/recall/mAP across 19 classes
  • Future work: Dockerisation, MLflow integration, CI/CD

🧭What I Focus On

  • Designing production-ready ML pipelines that balance quick experimentation with long-term scalability
  • Applying MLOps principles such as reproducibility, version control, and experiment tracking to improve workflow consistency
  • Framing proof-of-concepts in business terms, connecting technical insights to operational value and future deployment potential
  • Building AI capability from the ground up, including environment standards, governance frameworks, and best-practice workflows

📫 Let’s Connect

  • LinkedIn
  • Always open to discussing senior DS roles , or opportunities to drive AI maturity at scale.

Pinned Loading

  1. digitised-pid-mlops digitised-pid-mlops Public

    Python

  2. predictive_maintenance predictive_maintenance Public

    HTML