Jash-Piam Parekh

B.S. in Mathematics, minoring in Statistical Sciences

GPA: 3.78/4.0

Relevant Courses: Machine Learning, Data Structures, Advanced Language Systems and Theory, Object-Oriented Programming

• Enhanced knowledge retrieval using GraphRAG, increasing query resolution speed by 45%
• Integrated LLM software to auto-generate neural network insights, improving report generation efficiency by 60%
• Optimized a hybrid architecture using GraphRAG and LLM, leading to a 50% reduction in computational cost
• Fine-tuned company LLM models for industry-specific tasks, achieving a 35% improvement in prediction accuracy

• Created an application to iterate through company servers for leaked SSH keys and private information
• It Scanned Slack Channels, DEVOX code, and GitHub for hard coded information
• Optimized run-time performance for key finder app by 21%, allowing for quicker safety alerts
• Developed a Slack Bot using Slack API for real-time reports, improving response times by 34%

• Taught Lua, Python, and C# programming to over 10 students simultaneously in live virtual sessions
• Designed and led hands-on projects involving Arduino and robotics to teach core CS concepts like loops, conditionals, and hardware integration
• Developed tailored lesson plans for students of varying skill levels, focusing on problem-solving, and logical thinking
• Fostered collaborative learning environments where students built games, automated systems, and interactive devices

• Built the platform using React, Node.js, and Docker with a scalable backend architecture designed to support thousands of concurrent learners
• Developed a simulation tool enabling users to counteract real time threats like DDos and phishing in a virtual environment. (average response time of 200ms for user-interactions)
• Employed dimension reduction and random undersampling techniques to improve model performance by +10%

• Designed a system integrating Apple Watch heart data and real-time video analysis to monitor sleep patterns
• Built a pipeline to synchronize heart rate and camera data for comprehensive sleep tracking
• Integrated an LLM to provide direct feedback on sleep quality by comparing data with scientific research

• Created a web app using React and JavaScript to track dining hall occupancy using PyTorch and YOLO (You Only Look Once) for AI-based camera tracking
• Used OpenCV for image processing and integrated with public APIs for real-time data
• Optimized response time by 90%, updating every 0.5ms
• Won 2nd place in our school's Data Science Project competition

Languages

Technologies

Certifications