About
My name is Yash Chennawar, and I'm an aspiring ML Engineer passionate about AI, software development, and quantum computing. I study computer science at Rutgers University-New Brunswick Honors College. Currently, I'm a research assistant exploring brain-inspired neural networks and their applications in deep learning. I've built AI-driven apps, VR healthcare simulations, and machine learning solutions, always striving to innovate and learn.
Work Experience
Lockheed Martin - Space
As a Platform Executive, develop an ML pipeline on Azure & Databricks to automate contract data extraction for the leading corporate tax firm. Reduced costs by 99% after evaluating 3 methods: GPT Vision, OCR+GPT, and Azure Document Intelligence. Collaborated with data engineers and tax experts to enhance tax.com workflows with AI, significantly optimizing document processing while maintaining performance.
Research and develop neural networks inspired by biological brain mechanisms to advance pattern discrimination tasks as solutions for the deep learning synaptic credit assignment problem, and publishing a paper for Neuron. Leverage PyTorch in Python for building and fine-tuning neural network models for spiral and MNIST datasets. Conduct validation to optimize performance with HPC multithreading on Linux compute cluster with shell scripts. Began researching through Aresty Summer Science. Working with Dr. Aaron Milstein and PhD students.
Built multiple AI apps/websites covering diverse domains from pantry trackers to code assistants. Designed, developed, and deployed applications using MVC patterns, Agile, CI/CD, and microservices under the guidance of engineers from Amazon, Bloomberg, and Capital One.
Developed VR/MR applications for Apple Vision Pro and Meta Quest 2 & 3 to enhance medical education and practice. Collaborated with 2 cross-functional teams in a startup environment to design and implement immersive simulations, reducing medical errors and improving healthcare outcomes. Utilized Unity and C# to create interactive environments, ensuring high performance and user engagement. Presented project progress and outcomes to surgeons & the US military, receiving positive feedback and valuable insights.
Mentored students in robotics for the World Robotics Olympiad, First LEGO League, and Sumo Robot League events. Taught Python and Java, focusing on automation, fundamental engineering principles, and sensor control.
Education
Societies: Phi Beta Kappa Honor Society Coursework: Introduction to Deep Learning, Introduction to Artificial Intelligence, Introduction to Data Science, Computer Architecture, Data Structures, Principles of Information and Data Management, Design and Analysis of Computer Algorithms, Discrete Structures (I & II), Linear Algebra, Elementary Differential Equations, Calculus III, Probability Theory, Tensor Networks
Skills
Some cool stuff I've built
I've developed many projects, AI-powered web applications to machine learning models. Here are some of my favorites.
PodVibe.fm
Created agentic AI web application that generates key insights from videos/podcasts, letting users navigate to key concepts. Used Google Cloud's Gemini and YouTube Data APIs for autonomous analysis and content summarization.
SkillSync
Built AI platform for automated team role assignment and generating role specifications. Processes resumes and GitHub profiles using LLMs to match skills with project needs, enhancing team efficiency.
RUControversial
Developed an NLP-powered tool to predict audience reactions to Reddit posts before publishing, leveraging a fine-tuned DeBERTa-v3 transformer model for social sentiment prediction.
WeDormin?
Created a web app to help college students find roommates and housing. Users can create profiles, search for roommates, and find available dorms. Integrated AI to match users based on preferences.
Frantry
Developed iOS and Android mobile app to scan receipts with OCR, track pantry items, predict shelf life, and generate recipes with AI to reduce food waste.
MBS Dashboard
Developed a full-stack MBS management system to filter mortgages, calculate costs, manage pools, and add new entries, streamlining complex financial operations efficiently.
Binder
Developed a web app to generate personalized book recommendations similar to Tinder. Recommended books using TF-IDF and KNN.
Stock Prediction Model
Created an ML model to predict S&P 500 index values using historical stock open and close time series data from yfinance. Trained data with a Random Forest Classifier by using various values for each day.
Building things with friends
I love participating in hackathons and learning about wonderful and innovative ideas from my teammates and other groups.
- O
ODSC x Google Cloud Agentic AI Hackathon
Manhattan, New York
Developed PodVibe.fm: Agentic tool to generate key insights from videos/podcasts and fast forward to important segments. - K
Kaggle x Google Agents Intensive Capstone Project
Online
Developed CrisisNet: Agentic AI system for real-time disaster monitoring and actionable safety suggestions. - H
HackRU Fall 2025
New Brunswick, NJ
Developed SkillSync: A platform to automatically assign team roles and generate role specifications using LLMs. - B
Bank of America Code-A-Thon ~ Finalist
Charlotte, NC
Developed WeDormin?: A web app to help college students find roommates & housing, and get personalized roommate recommendations. We advanced to the finals at the BofA headquarters. - H
HackRU Spring 2025 ~ Winner
Piscataway, NJ
Developed Frantry: Mobile app to scan receipts, track pantry items, predict shelf life, and generate recipes with AI to reduce food waste. Won Wakefern's Overall Best Project Award. - H
HackRU Fall 2024
New Brunswick, NJ
Developed TripWhiz: AI-powered trip itinerary planner that optimizes travel routes and provides personalized guidance. - H
HackRU Spring 2024
Piscataway, NJ
Developed Binder: Web app to generate personalized book recommendations using ML algorithms.
Campus Involvement
Organizations I'm involved with at Rutgers University.
Developing ML mentor matching algorithm using word embeddings, semantic similarity, and Google Sheets integration. Leading a committee for technology and neuroscience projects. Hosting interactive workshops educating 100+ members.
Developed autonomous aircraft to complete waypoint navigation, image capture, target recognition, and air delivery. Used Python to create computer vision ML models for image recognition with convolutional neural networks.
Research and development of neural networks inspired by biological brain mechanisms to advance pattern discrimination tasks as solutions for the deep learning synaptic credit assignment problem.