Tianyang | Machine Learning Engineer

👋 About Me

I am a Machine Learning Engineer focused on building practical and scalable AI systems. My expertise spans the full lifecycle of machine learning projects, including data analysis, feature engineering, model training, evaluation, and deployment. I have hands-on experience in recommendation systems, predictive modeling, and intelligent decision-making systems.

🧠 Technical Skills

I specialize in Large Language Models (LLMs), Machine Learning, Deep Learning, and AI Infrastructure, with a strong focus on building production-ready, scalable AI systems.

Machine Learning

Supervised & Unsupervised Learning
Feature Engineering & Model Evaluation
Linear Models, Tree Models, Ensemble Methods

Deep Learning & LLMs

Neural Networks, CNN, RNN, Transformer
Large Language Models (Fine-tuning & Inference)
Prompt Engineering & Representation Learning

Retrieval & Recommendation

Collaborative Filtering & Ranking Models
Vector Databases & Similarity Search
Retrieval-Augmented Generation (RAG)

LLM System Architecture

Data Ingestion & Embedding Pipelines
RAG Workflows & Context Grounding
Evaluation, Logging & Monitoring

MLOps & AI Infrastructure

Model Deployment & Serving
Cloud-native AI Systems (Azure)
Docker, Kubernetes, CI/CD for ML

Data & Engineering

Python, NumPy, Pandas, SQL
Scalable Data Pipelines
Linux, Git, System Debugging

🚀 Projects

AI Recommendation System

Designed and implemented an end-to-end recommendation system, including user profiling, feature pipelines, ranking models, and offline evaluation. The system supports both offline training and online inference.

Tech: Python, TensorFlow, NumPy

Predictive Modeling

Built predictive models based on historical data for business forecasting and risk analysis, with a strong focus on model robustness and interpretability.

Tech: Scikit-learn, XGBoost

🧩 LLM System Architecture

I design and build end-to-end LLM-powered systems with a strong focus on scalability, reliability, and real-world applicability. A typical system architecture includes:

Data Ingestion & Processing: Structured and unstructured data pipelines for documents, logs, and user-generated content.
Embedding & Vector Storage: High-dimensional embedding generation and storage using vector databases for efficient similarity search.
Retrieval-Augmented Generation (RAG): Combining retrieved context with LLMs to produce accurate, grounded, and explainable responses.
Model Serving & Scaling: Containerized model deployment with autoscaling and monitoring.
Evaluation & Monitoring: Continuous evaluation, logging, and feedback loops for model quality and system performance.

Tian Yang