Divakar Verma

Software Engineer | AI | ML Inference | Computer Vision

United States
LinkedIn | GitHub | Email

My broad interests:

  • ML model Inference and Performance Optimisation
  • Computer Vision; Camera; Imaging

I carry 3 years of work experience shipping AI based sofware solutions. I’m enrolled as a Masters in Computer Science at the Rutgers University [Graduating in May 2023] and currently working as an AI Research Assistant at the Abraira Lab on a video classification project.

Last summer I interned at Nvidia as a Deep Learning Performance Engineer to develop a highly optimised Inference pipeline of RNN-T Speech model for MLPerf-Inference benchmark. Previously, I worked with Samsung R&D Institute and Flipart Inc. to ship Camera-Imaging and Speech solutions respectively.


Sep 2, 2022 Advanced Computer Vision at Princeton University; Prof. Jia and Prof. Olga.
May 20, 2022 Nvidia Internship - Deep Learning Inference Performance S/W Engineer
Jun 7, 2020 Global patent - A novel technique & application for slow motion based videos & images.
Nov 20, 2019 Paper - “Deep Demosaicing using ResNet-Bottlenect architecture”
Feb 20, 2019 Professional software coding competency across Samsung R&D centers.
Mar 12, 2018 Winner, ExxonMobil Hackathon - rule compliant logo detection & geotagging. PPT
Feb 20, 2017 Undergraduate research thesis at the University of Melbourne, Australia
Feb 20, 2014 Merit certificate in Mathematics.