HOME

Hi, welcome to my web page. 

Myself Ravi Theja, working as Data Scientist III at Inmobi - Glance group. 

NEW: 

I am an oper-source contributor to LlamaIndex. Check post1 and post2 on twitter about my work.

VideoQues: A GPT-3 and whisper based app built using gpt-index to answer queries on content in a video. Demo - here

DocsQues: A GPT-3 based app built using gpt-index to answer queries on a textual document. Demo - here.

FactGen: A GPT-3 based app to generate facts about entities present in the text. Demo - here.

My work primarily involves in improving user experience and engagement on Glance TV and Glance feed platforms. Broadly speaking I have leveraged skills in NLP, Deep Learning, and Recommender Systems to enhance user experience.

Prior to this, I worked as a Research Engineer at TCS Innovation Labs (TCS - TRDDC) where our work on humor recognition in edited news headlines got accepted to COLING-2020.

Prior to this I completed my Master's from IIIT-Bangalore with a specialization in Deep Learning, Natural Language Processing, and Computer Vision.

My master's thesis is titled "Generalised Natural Language Understanding (NLU) using Attention networks".

I am also the winner of the NIPS-2017 Paper Implementation Challenge. I stood 28th (top 2%) in Microsoft India AI Challenge 2018 and stood 5th in Hope Speech Detection shared-task in LT-EDI workshop at EACL-2021.

Prior to Master's  I worked as Data Scientist at a hyper growth startup in India where I worked on sales forecasting and text classification problems.

GitHub: https://github.com/ravi03071991

Medium blog: https://medium.com/@ravidesetty

I earned my Bachelors from National Institute Of Technology, Warangal (NIT-W) (July 2009 - April 2013).

I am currently looking for working opportunities in the area of Natural Language Processing(NLP), Deep Learning(DL), and Recommender Systems areas.