Compute and Memory Resource Estimation for ML Pipelines
Designed a metamodel for predicting ML models’ CPU/GPU & memory requirements, reducing cost by 25%.
I am passionate about using technology to solve complex engineering challenges and build systems that drive meaningful impact. As a motivated graduate student pursuing a Master of Science in Computer and Information Technology at Purdue University, I am dedicated to deepening my expertise in software development, cloud infrastructure, and machine learning. My recent experience at Amazon Web Services (AWS Bedrock) allowed me to contribute to the development of scalable systems for managing large-scale model deployments. At Razorpay, I played a key role in re-architecting real-time data services and building secure, high-performance backend systems. My internship at Jaguar Land Rover further strengthened my skills in machine learning pipeline development and predictive modeling. As a Graduate Research Assistant at Purdue, I focus on building reliable data pipelines and researching intelligent resource scaling for cloud-native ML workloads. These experiences have sharpened my proficiency in languages like Python, Golang, Scala, and Java, as well as technologies such as Spark, TensorFlow, Kubernetes, and Docker. Driven by a strong desire to learn and innovate, I am eager to apply my skills in impactful ways. Let’s connect—I’d love to explore how I can contribute to your team’s success.
Recent Work
Detection of Suicidal Ideation through Deep Learning
Developed a model to classify suicidal texts with an F1 score > 0.975 using BERT and Graph Sage.
Go Authentication Module for Trino Gateway
Enhanced security & access control for 2000+ users at Razorpay by programming a Go authentication module.
LookML to Cube.js Conversion Tool
Automated the conversion of LookML to Cube.js, reducing resource consumption by 70% at Razorpay’s hackathon.
Face Mask Detection with CNNs
Developed a real-time face mask detection system with 95% accuracy using CNNs and OpenCV.
Gender Prediction & Image Classification Web App
Built a Flask web app for gender prediction with 99% accuracy and image classification using SGD.
RESUME
COVER LETTER
Get In Touch
Whether you're interested in working together, have a project in mind, or just want to discuss the latest trends in technology, I’d love to hear from you. Reach out, and let's make something great happen!