Resume ​

Please also see my github.

Education details ​

Publications ​

[1] M. L. Maher, Y. Tadimalla, and D. Dhamani, “Is ChatGPT Good for Your Students? A Study Design of the Impact of AI Tools on the Student Experience in Learning Java” EDULEARN23 Proceedings, pp. 5702–5709, 2023, doi: 10.21125/edulearn.2023.1493.

Work Experience ​

• Adjunct Lecturer at UNC, Charlotte
  January 2020 - Present

  Currently teaching ITSC 1212 Introduction to Computer Science 1, ITSC 2214 Data Structures, and ITCS 3134 - Digital Image Processing at the College of Computing and Informatics at UNC, Charlotte. Would have taught 21 sections by the Fall of 2022.

• Machine Learning Consultant at Powerhouse Consulting
  February 2020 - April 2020

  Helped write two funding proposals, an SBIR from the Department of Education, and another for NSF on using a reinforcement learning-based intelligent aide for students with Autism spectrum disorder (ASD).

• Tutor at UNC, Charlotte
  January 2020 - May 2020

  Tutored for ITCS 3181 - Introduction to Computer Architecture for the Spring 2020 semester at the College of Computing and Informatics at UNC, Charlotte.

• Lowe's Innovation Labs
  August 2018 - December 2018

  Collaborated with Lowe's Innovation Labs for a single project of developing an MVP for a Visual Search system meant to identify products within Lowe's massive product catalog from images. (Not an internship)

• General Secretary of Students' Council
  August 2016 - August 2017

  Served as the General Secretary of the St. Francis Institute of Technology Students' Council (SFITSC) from 2016-2017.

Academic Projects ​

• Solution for SemEval Task 3: EmoContext
  Natural Language Processing

  Implemented a moderately successful solution (ranked 44th/241 as of 12/13/2018) for SemEval 2019 Task 3: EmoContext: Contextual Emotion Detection in Text., using a non-traditional Bi-LSTM architecture that used a combination of FastText and some Custom word embeddings, that performed as it did without requiring any significant preprocessing

  Project Link: https://github.com/DhruvDh/emocontext

• Semi-supervised Natural Language Understanding
  Natural Language Processing

  Used generative language models to solve the Natural language Understanding task of user intent identification.

  Project Link: https://github.com/DhruvDh/semisupervised_nlu

• Visual Search Prototype for Lowe’s
  Computer Vision

  Collaborated with Lowe’s Innovation Labs to build a visual search system, that given an image, identifies the product within it, using a very small, computationally efficient Convolutional Neural Network. Was trained to detect products across the category of Decorative Light Bulbs.

  MVP Demo: https://youtu.be/WmK4QiC-7Ag

• Robotic Arm Manipulation using Quaternion Neural Networks
  Intelligent Robotics

  Trained a Quaternion Neural Network to predict what positions the links in a robotic arm should be in for the end-effector to be placed in the desired position. Effectively did the same job as an inverse kinematics solver but with better computational efficiency.

  Project Link: https://dhruvdh.github.io/intelligent_robotics_project/

• Speech-To-Text
  Web Application

  A progressive web application that generates transcripts from audios of recorded discussions, meetings, etc., and then uses NLP to perform a rudimentary analysis on them. Used - Vue.js, Node.js, and the Google Cloud Platform.

  Project Demo: https://youtu.be/KgCdh2ru6jg

• Intelligent to-do list
  Web Application

  An intelligent to-do list web application for students that could "understand" the intention behind a to-do list item and then proceed to classify to-do items and organize them automatically. It is easier to show than to explain, please refer to the demo

  Project Demo: https://youtu.be/FLrzDlr9oZM

• Efficient Implementation of the N-Body problem on CPUs and GPUs
  Web Application

  Implementation of the N-body problem that is highly optimized for Intel’s Haswell micro-architecture, written in Rust. The GPU implementation leverages CUDA and is written using the Futhark Parallel Programming language.

• Intent Recognition from Textual User Utterances
  Natural Language Processing

  Explored the feasibility of using a hierarchy of classifiers to classify the intent of a user, from textual utterances. Using the crowd-sourced Snips NLU dataset of 13,784 user utterances, the approach was found to have an overall accuracy of 95.16% while only using rudimentary text classification techniques such as Logistic Regression

  Project Link: https://github.com/DhruvDh/ItForSociety2018

Other Work ​

• umm: A java build tool for novices, also a scriptable autograder.
  Project Link: https://github.com/DhruvDh/umm

• dip_app: A course website to demonstrate and give student feedback on Digital Image Processing assignments
  Project Link: https://dhruvdh.github.io/dip_app/

Technical Skills ​

• Programming Languages:
  Proficient in: Rust, Python, JavaScript, Bash/Shell, Java, Go.
  Can also work with: C, C++, C#, TypeScript

• Frameworks and Tools:
  Machine Learning: PyTorch, Pandas, NumPy, Numba, libtorch, TVM, Tensorflow, StreamLit.
  Web Frameworks: Express.js, actix-web, rocket.rs, Vue, React

• Other:
  WebAssembly, git, webpack, OpenMP, CUDA, Docker.