V S Abhinav Rahul Gandrakota

V S Abhinav Rahul Gandrakota

Senior Year Undergrad

BITS Pilani

About Me

Hi! I am V. S. Abhinav Rahul Gandrakota, a final year Electronics & Instrumentation undergraduate at BITS Pilani, Hyderabad. I am mainly interested in the field of data science, especially Natural Language Processing and Artificial Intelligence.

Previously, I worked with Dr. Ankur Bhattacharjee on using Deep Learning based algorithms for development and optimization of an efficient VRFB thermal management system. I am currently working with Dr. Aruna Malapati on getting pre-trained embeddings for code-mixed data.

I got my first opportunity to work in the Deep Learning domain as part of my summer internship at North Eastern Space Application Centre, which involved 3D Semantic Segmentation task performed on satellite generated images. After that, I knew that I wanted to pursue my professional career in this domain.

I’ve worked on many assignments and projects as part of the courses done under the data science minor. Some of them include Sign Language Translation, Audio Based Language Identiifcation and Alzheimer Helper.

I’ve also worked on Robotics projects in the past, first in 10th grade, and then in my first year of college. Although I don’t enjoy the theoretical aspect of electronics, I have always found working with actual hardware a lot of fun. I hope to get the opportunity to incorporate my knowledge of AI/ML along with Robotics in the future.

I have always loved the subject of Mathematics, particularly Probability & Statistics. That was what, despite being an Electronics major, initially pushed me to explore the field of data science.

In my free time, I like to read books, play football and work out. I’ve also recently started learning how to play the guitar. I hope I can learn how to play my favourite songs very soon!

Feel free to contact me if you’d like to know more about my work!

Interests
  • Natural Language Processing
  • Deep Learning
  • Artificial Intelligence
Education
  • B.E in Electronics & Instrumentation Engineering, 2019-2023

    BITS Pilani

  • Minor in Data Science, 2021-2023

    BITS Pilani

Interests

football1
Football

Played at U-15 club tournaments

reading
Reading

Stories, Novels, Inspirational, and Business related books

rubiks
Speedcubing

Mainly 3x3

Work Experience

 
 
 
 
 
North Eastern Space Applications Centre
ML Intern
May 2021 – Jul 2021 India

Worked on development of DL pipeline for 3D semantic segmentation task performed on drone and satellite generated images. Pipeline was deployed as an API using an Inference server to use this service in an application or web app for visualization of results. The project involved:

  • Choosing an Inference Server that could handle multiple frameworks, was scalable, could run multiple models simultaneously, had readily available documentation and was easy to work with while being fully open source. The ones tested were MLflow, BentoML, Triton and Cortex. Triton Inference Server turned out to be the most optimal for our requirements.
  • Using Tensorflow and Nvidia CUDA to work with Deeplab for object detection and YOLO for semantic segmentation. Used Nvidia Container Toolkit to utilize Triton GPU.
  • Hosting pre-trained models using Inference serving. Postman client API was used for testing the inference server.
  • Working on an image classification model using CNNs.
  • Building a hosted website. It involved:
    1. React.js based frontend where users could uplaod images.
    2. Node.js based backend where the models could run on the uploaded images and send back the output in the form of a JSON based response.
    3. Cloud based deployment of web service using AWS.