Snake eating contribution cells and revealing TERRYH
Build systems.
Study deeply.
Ship clearly.
I build practical machine learning projects with a focus on tabular modeling, feature engineering, model evaluation, and clean Python/C++ implementation.
My current goal is to become a Machine Learning Engineer who can connect data preparation, model training, validation, and deployment into reproducible engineering workflows.
Machine Learning → Feature Engineering → Model Evaluation Algorithms → Data Structures → Problem Solving Software Systems → Docker / Linux → Deployment Basics
- End-to-end ML pipelines
- Regression modeling and ensemble methods
- Cross-validation and error analysis
- C++ data structures and algorithms
- Docker, Linux, Git, and MLOps fundamentals
A local-first daily planner that combines task scheduling, execution tracking, analytics, and ML-based duration prediction in a Dockerized multi-service system.
Focus
- FastAPI service architecture
- Scheduling and task prioritization
- ML duration prediction and retraining
- MySQL and Docker Compose
- End-to-end testing and documentation
An uncertainty-aware bike-share forecasting and rebalancing decision system that turns station-level demand predictions into operational risk scores and capacity-safe transfer recommendations.
Focus
- Station-level demand forecasting
- Time-aware validation and uncertainty estimation
- Live GBFS ingestion and station reconciliation
- Constraint-aware rebalancing decisions
- FastAPI and React operator dashboard
An end-to-end machine learning web application for training, comparing, versioning, and serving Titanic survival models through REST APIs.
Focus
- Scikit-learn pipelines
- Model evaluation and GridSearchCV
- Flask REST API and Ajax
- Model versioning and batch prediction
- What-if analysis and feature importance
An end-to-end regression project for Kaggle House Prices, combining robust feature engineering, strict cross-validation, model ensembling, calibration, and reproducible MLOps tracking.
Focus
- Fold-local preprocessing and feature engineering
- Repeated stratified cross-validation
- Lasso, XGBoost, and Huber GBR ensemble
- Out-of-fold blending and model calibration
- MLflow tracking and reproducible submissions
A system design portfolio project focused on backend engineering, API design, and practical software architecture.
Focus
- System design practice
- Backend architecture
- API design
- Engineering documentation
- Portfolio project development
C++ implementations for data structures, algorithms, and competitive programming.
Focus
- Trees and graphs
- Dynamic programming
- String matching
- CLRS-style practice
- Clean C++ implementation
IoT distance measurement system using ESP8266 and ultrasonic sensing.
Focus
- Embedded C++
- Ultrasonic sensor
- Low-pass filtering
- Buzzer feedback
- Wi-Fi web display
Additional machine learning, deep learning, and engineering workflow projects.
Repositories
- deep-learning-labs — Deep learning projects and experiments
- machine-learning-labs — Machine learning practice and experiments
- Serendipity — Epiphany — Reusable AI-assisted engineering workflows
AI / ML Engineering ├── Data analysis, preprocessing and visualization ├── Feature engineering and dimensionality reduction ├── Machine learning, validation and ensemble methods ├── Deep learning, CNNs and transfer learning ├── NLP, transformers, embeddings and semantic search ├── LLM applications, RAG, agents and evaluation └── MLOps, APIs, Docker and model monitoring Computer Science Foundations ├── Data Structures & Algorithms ├── Linear Algebra & Discrete Mathematics └── Operating Systems & Computer Architecture Software Engineering ├── Git / GitHub, Linux and Docker ├── Backend development and system design ├── PostgreSQL and MySQL └── Testing and reproducible development
I believe strong machine learning engineers should understand not only how to use tools, but also the principles behind them.
My goal is to grow into an engineer who can combine programming ability, algorithmic thinking, mathematical foundations, and real-world ML system development.
Let’s build something useful.
If you would like to connect, discuss machine learning, algorithms, software engineering, or potential collaboration, feel free to reach out.