📍 San Jose, CA
🎓 Data Science @ Evergreen Valley College
I’m a Machine Learning engineer focused on building end-to-end, production-ready AI systems, not just models in notebooks.
I enjoy working across:
- ML model development (CV, classical ML, DL)
- Backend systems (FastAPI, APIs, inference pipelines)
- Deployment (Docker, AWS)
- Data systems (SQL, analytics, feature engineering)
Currently exploring RAG systems, LLM applications, and scalable ML infrastructure.
- Turning ML models into real applications
- Optimizing for latency, accuracy, and scalability trade-offs
- Designing clean, modular ML pipelines
- Shipping working systems, not just experiments
CNN (Keras) • OpenCV • FastAPI • Real-time Inference
- Built a real-time computer vision pipeline with webcam integration
- Achieved ~15 FPS inference with optimized preprocessing pipeline
- Designed modular FastAPI backend for inference serving
- Tuned trade-off between latency vs model accuracy
Scikit-learn • Streamlit • FastAPI • Docker • AWS EC2
- End-to-end ML pipeline: preprocessing → training → inference API
- Deployed containerized ML app on AWS EC2
- Integrated LLM (GroqCloud) for contextual prediction insights
- Built modular pipeline for easy retraining and scaling
PostgreSQL • Advanced SQL • Data Modeling
- Designed synthetic dataset (40K+ records) simulating hiring pipeline
- Implemented cohort analysis, window functions, and CTE-based queries
- Built analytics system to track funnel conversion behavior
- Focused on query optimization and relational modeling
TensorFlow • Deep Learning • Regression
- Built regression model for electrochemical performance prediction
- Improved training stability using normalization + dropout
- Designed reusable inference pipeline for deployment
- Evaluated model performance using error-based metrics
KNN • SciPy • Sparse Matrices
- Built collaborative filtering system using cosine similarity
- Optimized memory using sparse CSR matrix representation
- Implemented fast nearest-neighbor retrieval pipeline
- Retrieval-Augmented Generation (RAG) pipelines using vector databases
- Scalable ML APIs with FastAPI + Docker
- LLM-powered real-world applications
- Improving model deployment + monitoring workflows
This portfolio demonstrates:
- End-to-end ML system design (not just training models)
- Backend + deployment experience
- Applied machine learning across multiple domains
- Strong focus on productionization and scalability
To become a Machine Learning Engineer focused on production AI systems, building scalable and reliable ML infrastructure in real-world environments.