constcoder={name:'Viren Patil',skills:['React', 'Node.js', 'Express.js', 'Angular', 'Django', 'GCP', 'Express', 'MongoDB', 'Docker', 'AWS', 'PyTorch', 'Android Development'],hardWorker:true,quickLearner:true,problemSolver:true,hireable:function() {return(this.hardWorker&&this.problemSolver&&this.skills.length>=5);};};something about me ...
I’m Viren Patil, a software engineer who recently graduated with an MS in Computer Science from USC. I’m drawn to work that sits at the intersection of engineering and AI, and most of my projects reflect that. Over the past few years, I’ve built a collaborative movie watchlist app, an ephemeral encrypted chat platform, and several AI driven systems spanning large language models, multimodal video analysis, and automated sports commentary. I’ve also explored blockchain, data engineering, and cloud infrastructure, which has helped me develop a broad, practical toolkit for solving problems across different domains. What motivates me most is taking ambitious ideas and shaping them into thoughtful, reliable tools that people genuinely enjoy using.
2023 - 2025
Master's of Science in Computer Science
University of Southern California
GPA: 3.85/4.0
Coursework: Algorithms and Datastructures, Foundations of AI, Applied NLP, Web Technologies, Deep Learning, Machine Learning, Information Retreival and Web Search Engines
2018 - 2022
Bachelor of Technology - Computer Engineering
College of Engineering Pune
CGPA: 9.32/10.0
Coursework: Algorithms, Linear Algebra, DBMS, Computer Organization, Operating Systems, Object Oriented Programming, Web Technologies
2016 - 2018
Maharashtra State Board of Secondary and Higher Secondary Education
Pace Junior Science College Thane
Percentage: 94.15%
(September 2025 - Present)
Research Assistant
USC Viterbi School of Engineering
(July 2025 - October 2025)
Software Engineer - Backend Systems
Easley Dunn Productions
(June 2024 - August 2024)
AI Engineer
Radical AI
(July 2022 - June 2023)
Software Engineer
Citibank (Citicorp Services India Private Ltd.)
(May 2021 - July 2021)
Software Engineer Intern
Citibank (Citicorp Services India Private Ltd.)
Generating Commentary from Cricket Clips
constproject={title:'AI-Powered Commentary Generation for Cricket Clips',tools: ['PyTorch', 'Vision Transformer', 'LSTM', 'GPT-4', 'Streamlit', 'ElevenLabs TTS],myRole:Deep Learning Research & Full-Stack Development,Description: Built a multimodal deep learning system combining a Vision Transformer, LSTM, and GPT-4 to generate expert-level cricket commentary from video clips. Created a custom annotated dataset of 356 deliveries, predicted seven key match events, and integrated ElevenLabs TTS to produce real-time spoken commentary through a Streamlit interface:contentReference[oaicite:1]{index=1}.,};Movie Watchlist
constproject={title:'Personal Movie Watchlist Web Application',tools: ['React', 'Firebase', 'TMDb API', 'CSS],myRole:Full Stack Developer,Description: Developed a React + Firebase web app to track and share a personal movie watchlist. Features include scrapbook grid view, blurred-poster guessing mini-game, progress tracking, star ratings, and cloud sync across devices.,};GhostWhisp
constproject={title:'Ephemeral Encrypted Burner Chat Application',tools: ['React', 'FastAPI', 'WebSockets', 'ECDH', 'AES-GCM],myRole:Full Stack Developer,Description: Created a secure chat app with FastAPI WebSockets and React, featuring ECDH key exchange, AES-GCM encrypted messages, ephemeral sessions with graceful exits, emoji reactions, typing indicators, and image sharing. Built with a privacy-first design where no data is stored after sessions end.,};Open-Source LLMs Code Generation
constproject={title:'Evaluating Exhaustiveness of Synthetic Code Generated by Open Source LLMs',tools: ['Gemini API', 'Llama2', 'Chat GPT', 'Python', 'Java', 'C++', 'Javascript],myRole:Analyzing EvalPlus Framework & Generation of Seed Inputs,Description: Conducted a comparative analysis of Gemini and Llama-2 models for code generation tasks, demonstrating Gemini's superior performance across multiple programming languages. Highlighted Python's exceptional results due to its high-level syntax and rich ecosystem, underscoring the impact of model architecture and training methodologies.,};Localizing Moments in Surveillance Videos
constproject={title:'Localizing Moments in Surveillance Videos using Natural Language Queries',tools: ['NLP', 'Deep Learning', 'Python],myRole:Transformer Coding & Front-end management,Description: Developed a multi-modal system for smart city security that localizes moments in long CCTV video streams based on natural language queries. Introduced a novel dataset, "CamPark-Captions," and built a large-scale multimodal neural network architecture to enhance real-time crime detection and prevention. The system automates surveillance video screening, streamlining tasks that are otherwise manually intensive, with rigorous robustness testing to handle varying inputs.,};Stock Trading App
constproject={title:'Stock Trading Web Application',tools: ['Angular', 'Bootstrap', 'Express', 'TypeScript', 'MongoDB],myRole:Full Stack Developer,Description: A stock trading web application that lets you search stocks, add favorites, and build a portfolio by buying or selling stocks.,};Blockchain-Based Crowdfunding Application
constproject={title:'Blockchain-Based Crowdfunding Web Application',tools: ['Solidity', 'Ganache', 'Django', 'Bootstrap],myRole:Full Stack Developer,Description: Developed a decentralized crowdfunding web application using blockchain technology to ensure transparency, security, and guaranteed returns for backers. The platform allows backers to track project progress in real-time, ensures refunds if projects are aborted, and provides seamless interactions between admins, backers, and start-ups. Smart contracts automate transactions, eliminating third-party involvement and enhancing trust.,};© Content by Viren Patil