• Developed interactive charts for the web portal, allowing clients to easily and effectively track their inventory across different containers, locations, and timeframes using the Chart.js library

• Currently working on building a dynamic admin portal to provide clients with a way to configure their settings and monitor activity

• Developed and implemented an incremental data loading process from a PostgreSQL database to Databricks, using Spark and hash-based chunking to efficiently handle large datasets while minimizing the memory overhead

• Automated data transformation workflows by designing and developing Python scripts that leverage PySpark for ETL operations, ensuring optimized data processing and storage within Databricks.

• Created a chatbot to provide a natural language interface to search complex technical manuals written in Latex. Implemented a Retrieval-Augmented Generation (RAG) approach, integrating Open AI’s GPT-4 to generate responses based on user queries, strictly using the retrieved content while retaining a memory of previous questions and answers.

• Evaluated LLAMA3, OpenAI, and Gemini LLMs, testing different chunking schemes, embedding models, and prompts to optimize searching of complex technical content and statistical equations in the manuals.

• Used Python, with Streamlit and Langchain, along with Azure OpenAI GPT-4 and its text embedding models to develop the chatbot to capture user questions and render Latex output returned by the RAG model.

Teach students game of chess using the Steps method with weekly lessons. Motivate and encourage the students to participate in local competitions to get a United States Chess Federation official rating.

Interned at JOGO Health to help test hardware and software for their cutting-edge technology leveraging neuroplasticity to help patients with chronic conditions affecting movement.