Jason Ja

Developed an AI platform with FastAPI and OpenAI to automate document classification and summarization, reducing manual effort and improving turnaround time. Built a RAG-based semantic search using embeddings and Pinecone, implemented async processing with Celery and Redis for high-throughput inference, and deployed secure AI endpoints on AWS Lambda/API Gateway/S3.

Built a Python/Flask backend exposing REST APIs for ML-driven recommendation insights from raw user activity data. Created data preprocessing pipelines with Pandas and NumPy and improved prediction accuracy using scikit-learn classification models connected to frontend dashboards.

Created React.js dashboards to visualize ML predictions and analytics so users could better interpret model outputs. Optimized analytics performance by redesigning MySQL schemas and queries, adding indexing and refining query structures for high-frequency operations.

Developed Python analysis and visualization tooling using SciPy and Matplotlib to extract insights from large datasets. Built TensorFlow machine learning models for classification and trend prediction, created batch training/evaluation pipelines, and deployed models via Django-based APIs with unittest coverage.

Designed normalized PostgreSQL schemas to improve data integrity and reduce inconsistent/duplicated records across systems. Wrote advanced SQL queries and stored procedures for reporting, supported data migration, and improved slow query performance through indexing and query optimization.

Completed a bachelor's degree in computer science at California Polytechnic State University, San Luis Obispo from 2017 to 2020.