Aditya Ballaki - Portfolio

Education

Cornell University

Master of Engineering in Electrical and Computer Engineering

May 2025

Coursework in Machine Learning and Computer Science

Nanyang Technological University

Bachelor of Engineering in Electrical and Electronic Engineering

Jun 2024

NTU President Research Scholar

Courses in Computer Engineering

Skills

Programming Languages

Python Java C++ JavaScript / TypeScript CUDA Swift C# SQL MongoDB

Software Engineering

Flask Django Vert.x Kafka React ExpressJS NodeJS Qt Docker Kubernetes AWS

Machine Learning & AI

PyTorch TensorFlow LangChain OpenCV Keras Scikit-learn ElasticSearch

Experience

Aug 2024 - Present

Break Through Tech

AI Studio Course Support, New York, NY

Coached and mentored over 30 women in AI, guiding them through the development of their AI studio projects.
Fine-tuned Large Language Models and developed multimodal RAG chatbots for 6 custom datasets.
Assisted in building an object detection model with 95% accuracy using OpenCV, Python, Flask, and Docker.
Led training sessions on MLOps and Data Engineering.

Jun 2022 - Dec 2022

Zendesk

Software Engineering Intern, Singapore

Developed the backend for Zendesk Chat and Messaging, serving 200,000+ businesses daily.
Engineered a Python microservice for recovering 100% of failed messages in Kafka queues.
Resolved bugs and bottlenecks, reducing chat processing time by 10 minutes.
Refactored microservices with Vert.x Java and improved observability with metrics and logging.

Projects

Apple-Py-Torch

Python, CUDA, PyTorch

Built a machine learning library from scratch with tensors, automatic differentiation, and backpropagation. Achieved 5x speedup in matrix operations using CUDA and created a custom CNN with 90%+ accuracy.

Brew

React Native, NodeJS, AWS, MongoDB

Developed a university networking app using stable matching algorithms and reinforcement learning. Built with real-time chat functionality and tested successfully with 50 live users.

OrbitShield

Python, C++, Flask, TensorFlow, PyTorch, Linux

Developed an ML solution to protect space electronics from SEL using an LSTM neural network on a Raspberry Pi Zero with a 300ms response time.

Hello, I'm Aditya Ballaki