Help & Supportcareers@gravityer.com
Accelerating machine learning @Nvidia | LLM data curation , inference.
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
7+ years of making things go faster on GPUS at scale. Working on the NeMo Curator, NV-Ingest, https://rapids.ai/ for Nvidia's AI Infrastructure. Experienced with accelerating Distributed Computing, Data Science, Graph Neural Networks, Large Language Models. Expertise includes optimizing terabyte-scale data curation for foundation model training, model inference
The Johns Hopkins University
Master of Science - MS, Computer Science
January 1, 2017 – January 1, 2019
Delhi College of Engineering
Bachelor of Technology (BTech), Mathematics and Computer Science
January 1, 2012 – January 1, 2016
NVIDIA
Senior Software Engineer (Machine Learning/Data Science)
July 1, 2021 – Present
NVIDIA
Software Engineer (Machine Learning/Data Science)
February 1, 2019 – July 1, 2021
Expedia Group
Software Engineer
May 1, 2018 – August 1, 2018
San Francisco, California
The Johns Hopkins University
Machine Learning Researcher
January 1, 2018 – January 1, 2019
The Johns Hopkins University
Course Assistant
January 1, 2018 – December 1, 2018
The Johns Hopkins University
Research Assistant, Centre for Language and Speech Processing
January 1, 2018 – September 1, 2018
Citi
Software Engineer
July 1, 2016 – August 1, 2017
Pune, Maharashtra, India
Cvent
Software Engineer
June 1, 2015 – August 1, 2015
Gurgaon, India
Enactus DTU
Head of Project
August 1, 2012 – July 1, 2016
Parallelizing AdaBoost on Multi Core Machines (C++/ OpenMP)
March 1, 2018 – May 1, 2018
• Created a parallel implementation of Adaboost with a 22.14x speed up over a serial implementation using OpenMP
Visual Question Answering (Pytorch/Python)
February 1, 2018 – July 1, 2018
• Designed a model than can answer abstract questions based on images • Implemented Stacked-Attention and various LSTM+CNN models
Face Similarity Detection using Siamese convolutional neural networks (PyTorch)
November 1, 2017 – Present
Implemented a Siamese convolutional neural networks for face similarity detection using Pytroch
Classification, localization and tracking progression of Thoracic Diseases Using Deep Learning
August 1, 2017 – January 1, 2018
• Designed a model that can classify and localize thoracic diseases using a dataset of 112,120 frontal Xray images. • Model created by using transfer learning on Resnet50. • Disease Progression tracking over follow-ups using activated heat maps obtained from convolution layer
Classification of Arrhythmia Using Conjunction of Machine Learning Techniques
August 1, 2015 – April 1, 2016
-Research paper presented at presented at 15th International Conference on applied computer and applied computational science (ACACOS’16) organized in Prague, Czech Republic. -Applied rigorous data pre-processing and feature selection techniques and used machine learning techniques such as Neural networks, Decision trees, Random Forests, Gradient Boosting and Support Vector Machines. Used conjunction of the above algorithms to achieve maximum accuracy till date of 84.32% Secured 2nd position in “Technical Paper Presentation” during Cogenesis held at DTU. Feb 2016
Content Management System
June 1, 2015 – July 1, 2015
End to Design and Development of a content management system for managing slideshows across multiple displays Server Side Scripting done in Ruby on Rails Client Side Scripting done in React-Js Functionality for real-time update of content across clients using web sockets Tested and deployed successfully on company server for usage across departments
A Sentiment and Interest Based Approach for Product Recommendation System
December 1, 2014 – March 1, 2015
Research paper presented at Emmanuel College, Cambridge University in IEEE UK-SIM 17th International Conference on Modelling and Simulation Recommendation system based on computing correlation between entities using Interest Graph in conjunction with Sentiment Analysis Customized to calculate public sentiment of a brand Can be used for recommendations ranging from whom-to-follow on Twitter to online buying
Computer Vision for Hand-Writing Recognition
June 1, 2014 – Present
Digit recognition system based on one-vs-all logistic regression and neural networks Recognizes hand-written digits with an accuracy of 95.3%+; Developed using dataset of 5000+ examples using Octave
Project Unmoolan
September 1, 2012 – May 1, 2016
-Set-up a self-sustainable community water purification plant for a slum in Pitampura, Delhi -300+ people given access to clean drinking water at a cost less than Rs 30 per month. -Created empowered self-help groups of 4+ women to improve health conditions in slums -60,000+ impacted directly through health camps, awareness campaigns and other initiatives -In-process of rolling out a teaching app to aid in the elementary education of the slum kids
R Programming A-Z™: R For Data Science With Real Exercises!
Udemy
June 24, 2026 – Present
Online Course -Statement of Accomplishment-Machine Learning
Stanford University
June 24, 2026 – Present
Cultural Fit Analysis
The candidate's diverse project portfolio, including social impact projects like 'Project Unmoolan', indicates a well-rounded individual with interests beyond pure technical work. Their academic and professional trajectory shows a consistent focus on advanced ML and distributed systems, aligning well with a high-performance, research-oriented ML engineering culture. The breadth of skills from full-stack development (Ruby on Rails, React-Js) to deep learning research suggests adaptability and a willingness to tackle varied challenges.
Soft Skills & Operational Fit
The candidate's project descriptions, especially 'Project Unmoolan', suggest strong initiative, leadership, and a commitment to impactful work. Their role as a Course Assistant at Johns Hopkins also indicates an ability to communicate complex technical concepts. The experience at NVIDIA in accelerating data processing and machine learning using GPUs at scale points to a strong operational fit for performance-critical ML engineering roles.