Sanjay Karunamoorthy
If the ChatBot tells you something went wrong, please email me at sanjaykarun@gmail.com to remind me to take my database (AstraDB) out of hibernation. Thanks!
Ask the chatbot anything you want about me
Click the little icon in the top bar to activate the AI chat.
About Me
Who am I?
I'm a 4th year at UVA pursuing a B.S in Computer Science, B.A in Mathmatics, and Minor in Data Science expected to be completed by Spring of 2026. I am passionate about solving problems and learning anything cool.
Skills
Languages
- Java
- Python
- C
- JavaScript
- Typescript
- HTML + CSS
- SQL
Technologies/Frameworks
- Spring
- Django
- React
- OpenSearch
- Next.js
- Node.js
- PyTorch
- MongoDB
- Express
- Kafka
- Git
- REST API
Professional Experience
University of Virginia
Undergraduate Artificial Intelligence Researcher
August 2025 – Present
- Training Reinforcement Learning Agents to collaborate with humans to improve task performance with Proximal Priority Optimization. Conducting experiments with multi-agent settings to analyze cooperation, incentives, and emergent behaviors.
GEICO
Software Development Intern
June 2025 – August 2025
- Pioneered an MVP dashboard for a Project Development Life Cycle management system based on Shift Left methodology; created a REST API leveraging Django, PostgreSQL, TypeScript, and React; scoped to benefit 3000+ engineers.
- Partnered with 15+ Senior Directors, VPs, PMs, Engineers and SMEs in a cross-functional team to design a product workflow optimization system with JSON integration, Redis caching, GraphQL querying, and Azure DevOps/Slack automation.
- Implemented efficient GET/POST requests for bidirectional synchronization and timed system reloads to prevent race conditions.
- Engineered an email automation script by pre-processing 17,000+ repositories to reach out to 2000+ system owners; utilized Power BI and Excel for data input organization and communication for stakeholders, using pandas for email sending automation.
3M
Software Engineer Intern (Contract)
May 2024 – July 2024
- Crafted an efficient data processing solution leveraging Spring Boot and Spring Batch, handling over 2,000,000+ search data points belonging to 3M from Salesforce and Sharepoint, with thorough error handling.
- Optimized and utilized scalable big data pipelines with Apache Airflow and Kafka, enabling real-time streaming data integration with batch processing, leading to a 6% improvement in processing efficiency for semantic search, based on time and CPU usage.
- Utilized Amazon Bedrock for Natural Language Processing (NLP), employing Titan Text Embeddings to transform words into high dimensional vectors, enabling vectorized search for Retrieval-Augmented Generation (RAG).
- Incorporated kNN vectors (k=5) for embedding storage with lambda functions on OpenSearch, prompt engineering Anthropic Claude API to generate context-relevant responses; leveraging Swagger for API development.
SS Technology
Full Stack Engineer Intern
December 2023 – May 2024
- Directed a group of interns to develop a REST API, organized meetings via JIRA for agile sprint management. Worked on CI/CD pipelines for an employee management app using Docker, Kubernetes, Jenkins + Bitbucket, reducing deployment times by 30%.
- Designed and created a secure back-end authentication system using React, Express, Node.js, and MongoDB, utilizing JWT and Bcrypt to enhance security by eliminating all major authentication vulnerabilities by replacing an insecure SHA-256 hash.
- Automated monitoring tasks through Linux/Unix environments with Bash scripting to reduce manual setup time.
My Projects and Activities
Micrograd
- Implementing backpropagation over a dynamically built DAG. Creating a neural network from scratch.
TensorFlow, Python, Neural Networks, Backpropagation, Gradient Descent
August 2025 – Present
SK-GPT
- Implemented ChatGPT from scratch using neural networks with custom backpropagation and gradient descent algorithms.
TensorFlow, Python, Neural Networks
August 2025 – Present
Siblings or Dating – Facial Recognition
- Created a Convolutional Neural Network (MobileNetV2) using transfer learning to classify preprocessed images scraped from Reddit via PRAW.
- Achieved 80% validation accuracy distinguishing couples as either siblings or dating through hyperparameter tuning.
TensorFlow, Keras, Python, CNN, Django
March 2025
Akuna Capital Quant Trading Challenge
- Pioneered a python-based market making quantitative trading framework integrating nonlinear optimization and stochastic modeling.
- Implemented reinforcement learning to dynamically synthesize adaptive market intelligence.
Python, Reinforcement Learning, Nonlinear Optimization
August 2024
Poker Bot
- Used a Monte Carlo counterfactual regret minimization algorithm to approximate a Nash Equillibrium strategy for a 2-player Khun's poker game.
Python
June 2024 – July 2024
TenantTalk
- Built a web app for rating landlords and property management, adhering to Scrum methodology with a team of 5 other developers through sprints and weekly meetings.
Django, PostgreSQL, Heroku, AWS S3
Jan 2024 – Present
Coccidiosis Classification
- Constructed a Convolutional Neural Network (CNN) to detect Coccidiosis signs in chicken images, with the use of over 1000 labeled images.
Python, TensorFlow, Azure
Jan 2024 – March 2024
Film Finder
- Created using Python and Jupyter Notebook with libraries like Numpy, sci-kit-learn, pickle, and requests.
- Integrated TMDB API for data retrieval, including plot summaries, genres, and ratings from over 5000 movies.
Python, TensorFlow
Sept 2023 – Dec 2023
EzCite
- Lead developer of the website, utilizing React and Next.js.
- Leveraged a mix of the Perplexity LLM and Node.js modules such as discord.js and Puppeteer.
- Use of citation-machine to create optimal citations from an input of text.
React, Next.js, Node.js, JavaScript, HTML, CSS
Mar 2024 – Mar 2024
Natural Language Processing Application
- Detects names, cities, states, and countries from given sentences.
- End-to-end development of a robust NER application leveraging the Stanford CoreNLP.
Java, Spring Boot
Nov 2023 – Dec 2023
European Option Pricing
- Developed a Monte Carlo simulation-based pricing model for European vanilla options in C++using risk-neutral pricing, Geometric Brownian Motion, and the Box-Muller algorithm, achievingresults closely matching Black-Scholes solutions.
C++
May 2024 – June 2024
EDO.ai
- Developing and researching models to predict diabetes using extensive blood test data and other methods.
PyTorch
August 2024 – Present