Software Engineer @ NVIDIA
About Me
I'm a software engineer working on CUDA safety at NVIDIA, where I spend my time building systems for AI running on safety-critical platforms. I enjoy taking ideas from rough prototypes to something that actually runs in production, and collaborating with people across the whole build process.
I'm a big believer in using Gen AI as a force multiplier while building software. Tools like Cursor, Claude Code, and Codex are a natural part of my workflow, and I enjoy figuring out how to combine AI-assisted speed with solid engineering.
Outside of work you'll probably find me playing sports, travelling somewhere new, or out on a trek. I tend to thrive in collaborative and exploratory environments.
Experience
Where I've been building, learning, and occasionally breaking things.
Software Engineer — CUDA Safety Team
Joined as a SPRING Intern on the OpenCL Driver team and transitioned to full-time on the CUDA Safety team. Building projects for safety-critical autonomous driving platforms.
Research Engineer — Aerospace & Defence
Engineered deep learning cost estimation framework with Explainable AI dashboards for manufacturing process optimization.
Data Analyst Intern
Built ETL pipeline on 150K rows integrating Magento and Viniculum APIs. Deployed unified analytics store on Zoho for real-time trend analysis.
Research and Publications
Fun weekend rabbit holes that turned into something interesting.
Prediction of Post-Stroke Vascular Events: A Novel Two-Stage Machine Learning Model Integrating Platelet Function and Clinical Data
Had the opportunity to work with multiple faculty members of PES University and doctors to develop a novel two-stage ML model integrating biomarkers and clinical variables to predict recurrent vascular events after ischemic stroke. Uses Random Forest regression and classification with GAN-based class balancing.
Explainable AI: A Retrieval-Augmented Generation Based Framework for Model Interpretability
Introduces a novel RAG-based framework leveraging Large Language Models and domain-specific knowledge bases to provide clear, interactive explanations of model outputs. Demonstrated across healthcare, finance, and manufacturing, offering a scalable solution for making ML/DL systems more trustworthy and accessible for non-technical users.
View PaperYouTube Viralomics: A Scoring Framework by Analyzing Engagement Metrics in the Indian Context
Develops a unique 'viralomics' framework to measure virality using 10,000 trending YouTube videos from Indian creators. Employs statistical techniques for feature elimination, regression analysis, and Particle Swarm Optimization to derive a virality score achieving R² of 0.913 and MSE of 0.0020, validated through a two-step process.
View PaperEducation
Four years learning how computers think — and then AI showed up and started doing it faster
PES University
B.Tech Computer Science Engineering
- MRD Scholarship — Awarded to the top 20% of students based on academic performance.
- Teaching Assistant — Centre of Innovation and Entrepreneurship. Mentored student teams and supported entrepreneurship programs focused on product development and innovation.
- Entrepreneurship & Innovation — Collaboratively led a team through ideation, prototyping, and customer discovery to build an Extended Reality educational platform, developing and validating a Minimum Viable Product (MVP) through user feedback.
- Leadership & Events — Event Lead for E-Summit and Maaya (2023, 2024), organizing entrepreneurship and innovation events and interacting with founders, investors, and student builders. Also actively participated in hackathons and coding competitions across campus.
- Coursework — OS, DSA, Compiler Design, Databases, Cloud Computing, ML/AI, Big Data (Kafka, Spark).
Technologies
Tools and technologies I work with day-to-day.
Technologies
CUDALanguages
Python
C
C++
JavaFrameworks & Libraries
Django
Node.js
Next.js
Keras
PyTorch
TensorFlow
Scikit-LearnDatabases
MongoDB
MySQLTools
Git
Docker
Kubernetes
Jenkins
Claude CodePlatforms
LinuxProjects
CloudCommerce
Microservices e-commerce platform with REST APIs, Docker/Kubernetes orchestration, and automated CI/CD via Jenkins.
AttendAI
Real-time facial recognition attendance system for educational institutions using OpenCV with MySQL backend.