Ashish Sharma

Implemented in-memory caching with TTL (Time to Live) to optimize SQL database query performance, reducing retrieval times and minimizing database load. Implemented swarm agents to optimize response generation by dividing tasks across multiple agents, each specializing in specific domains, leading to faster and more accurate results. Implemented a background scheduler to run specific queries asynchronously, ensuring that users don’t have to wait for query execution and improving overall application performance.

Implemented a math graph/diagram generator using LLM (Large Language Model) and Matplotlib, enabling the automatic creation of mathematical visualizations based on textual input. Created an automated resume screening tool that helps recruiters evaluate resumes based on their alignment with job descriptions. The system automatically processes resumes uploaded to Google Drive, generates a CSV report with a score (out of 10), and highlights key points to streamline the recruitment process.

Developed a Router query engine that identifies and selects the appropriate tool based on user queries, utilizing Ollama’s LLaMA 3 as the large language model. Implemented RAG for various file types, including .pst files, audio files, messages, and SQL files, to enhance data retrieval and analysis efficiency significantly. Leveraged GCP to strategically implement the RAG system, significantly optimizing overall performance and ensuring robust data management and accessibility.

Coursework: Computational Physics, Image processing, Data Structure and Algorithm, Intro to ML