Sai Navya Jyesta

About Sai Navya Jyesta

I’ve always been fascinated by numbers and data – how hidden patterns emerge from complex datasets and how these insights can drive impactful decisions. This passion has guided me toward a career in data science, where I get to combine my love for problem-solving with technical expertise to make a meaningful difference. Currently, I’m pursuing my Master of Science in Computer Science at Indiana University Bloomington, with a focus on data science. I’ve worked on diverse projects, from analyzing global suicide rates to exploring microbial genome research. Using tools like Python, R, and TensorFlow, I uncover valuable patterns and trends that tell compelling stories through data. Whether through machine learning models or statistical analysis, I thrive on transforming raw data into actionable insights. Beyond my data science focus, I also enjoy building scalable, efficient systems as a full-stack and backend developer. From creating web applications with React, NodeJS, Spring Boot, and MongoDB to optimizing backend services, I always aim to deliver seamless, impactful solutions. I’m excited to take on new challenges in data science—discovering hidden insights and making sense of complex datasets—while also exploring opportunities in full-stack and backend development. Let’s connect if you\'re interested in collaborating or discussing opportunities in data science, full-stack development, and innovation!

Experience

1. Applied Ml Researcher - Data to Insight Center (D2i)

   Indiana University Luddy School Of Informatics, Computing, And Engineering

   Nov 2024 — Present

   Designed a state-of-the-art multimodal Retrieval-Augmented Generation (RAG) system using Paligemma Vision and ColBERT late interaction, generating image embeddings for textbooks and achieving a 170x speedup in semantic search over traditional BERT-based models- Leveraged Qdrant multi-vector store for high-speed retrieval, reducing query time to sub-100ms, and integrated LLaMA 3.2 Vision for context-rich responses, enabling 128K token cross-chapter exploration- Developed an efficient image-text retrieval system, significantly improving search accuracy and scalability- Currently working on optimizing metadata processing, testing retrieval techniques on diverse datasets, and refining similarity search and performance metrics.

Education

• Vellore Institute of Technology

  Bachelor's degree, Computer Science

  2018 — 2022

• Indiana University Bloomington

  Master of Science - MS, Computer Science