Jan 2021 – May 2021 · 4 mos

• Worked with DSAC lab (IIIT H), in collaboration with JP Morgan AI Research org, on actionable text extraction from SEBI documents
• PI - Dr Kamalakar Karlapalem

Aug 2020 – Dec 2020 · 4 mos

• Course on database and its applications, taken by 200+ undergrads
• Mentoring course projects, setting and grading quizzes and assignments, and conducting tutorials

Jan 2020 – May 2020 · 4 mos

• Head Teaching Assistant for the Graduate level, Advanced NLP course (covers Deep Learning + NLP) taken by PhD, Masters and Undergraduate (senior/junior) students.
• Responsibilities included setting & grading assignments, grading of exam papers, hosting tutorials and mentoring course projects, mostly related to Question Answering & Machine Translation.

May 2018 – Jun 2021 · 3 yrs 1 mo

• Worked under the guidance of Dr. Manish Shrivastava in Natural Language Processing (NLP) lab, Language Technologies Research Centre (LTRC), KCIS.
• Visual Question Answering (VQA)
• Information extraction from Legal Documents
• Detection and Annotation of Events in (monolingual) Kannada and code mixed Kannada - English
• Information extraction tasks of Named Entity Recognition and Event Detection in Kannada-English Code-Mixed Social Media data
• Published multiple papers during the same in A* and A conferences - COLING, LREC and RANLP.

May 2020 – Aug 2020 · 3 mos · Bengaluru, Karnataka, India

• Flagship Data Science (Consumer) team - LinkedIn Engineering
• To tag and start recommending events through My Network tab from the time of creation, I developed models using unsupervised ML with GSDMM & LDA + word embedding for labeling the events and analysed the results.
• Drove a 14% lift in invite acceptance rate. Preliminary analysis with LinkedIn’s Interest Graph was also done for 100M MAU.
• Set up a UMP flow to calculate CTR of event live video notifications on a daily basis and analysed the results
• Used: Python, scikit-learn, gensim, Hive/SQL

Aug 2019 – Nov 2019 · 3 mos · Hyderabad Area, India

• To automate the loan approval process and enable banks to analyse small transactions, built MVPs for the pilot tests using Django, MySQL & Docker, and developed predictive models using Statistical Machine Learning (ML) algorithms with careful feature engineering.
• Results from the pilot test for loan approval automation showed that the model was better than human loan officers where our system was able to lend 26.6% more at loan officers' delinquency rate or achieve 21.3% lower delinquency rate at loan officers' approval rate without compromising equity, led to a journal publication.
• Used: Python, Docker, Django, MySQL/SQL, scikit-learn, XGBoost
• I worked with Professor Prasanna Tantri, who is the executive director of Centre for Analytical Finance (CAF), ISB.
• Journal publication that resulted from the work - https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3518601

May 2019 – Jun 2019 · 1 mo · Hyderabad Area, India

• Worked on identification of cancer in liver tissue Whole Slide Images (WSI) and segmentation of affected areas using Deep Learning algorithms for Computer Vision.
• Used: Python, PyTorch, scikit-learn
• Onward Assist is an advanced cancer prognostics platform helping Oncologists take better treatment decisions and hence improve outcomes. It is part of the NASSCOM DeepTech Club, and part of Yale's Sustainable Health Initiative (SHI).

Aug 2017 – Nov 2017 · 3 mos · Hyderabad Area, India

• Led the SWE intern group of 3
• Built products to help users fetch and update details of their products and the track their inventory across different platforms.
• Designed and built a mobile application for android, and a webapp & a RESTful API using Django, React-JS and MySQL.
• Custom QR codes could also be generated for new products.
• Used: Java [Android app development], Django, MySQL, Python, React-JS (Javascript), HTML5, CSS3
• Marketfront Pvt. Ltd. is a SaaS-based enterprise solution for online and offline enterprise retailers to automate and help with their everyday functions.