Gaurav Sharma

Spearheaded a Google-led project aimed at predictive maintenance for switchgear in data centers, contributing to a 98% accuracy rate. Developed advanced algorithms to calculate health scores and optimize maintenance schedules. Designed and implemented a customized Looker dashboard tailored to address specific business needs, resulting in a 30% increase in data accessibility and a 20% improvement in decision-making efficiency. Played a pivotal role in the development of the Kreate Bot product, leveraging Retrieval Augmented Generation models (RAG) and fine-tuning techniques. Resulted in a 50% reduction in bot creation time and a 40% increase in user satisfaction through enhanced customization capabilities.

Generated synthetic data, increasing model performance by 25% and augmenting dataset by 50%. Spearheaded the implementation of an innovative text-to-image search system using Pinecone, an advanced vector database, resulting in a 40% reduction in search time. Leveraged NLP and transformer models for Q&A, achieving a 30% improvement in response accuracy through fine-tuning. Developed GAN architecture employing a generator and discriminator neural network, resulting in the production of synthetic data with a 95% resemblance to real data.

Relevant Coursework: Data Structures and Algorithms (C++), Prob & Stat in CS (Python), Intro to CS II (C++), Linear Algebra w/Computational Applications (Python)& D.B.M.S(Database Management System), Machine Learning & Artificial Intelligence,Compiler Design.