Software Engineer

• Mar 2023 - Present

Graduate Research Assistant - Behavioral Informatics Lab

• Sep 2021 - Jan 2023

- Identified bias in covid-19 autocomplete results across different languages using pandas, numpy, AWS S3 and Sagemaker.- Analyzed trends in google activity of users to understand the impact of covid-19 on their mental health using GCP and other tools.

Graduate Research Assistant - Sequence Analysis and Modelling Lab (SEQAM)

• May 2021 - Aug 2021

- Created Cross Modal Embeddings using images and text belonging to the same recipe. Increased performance by making use of vision transformers.- Boosted interpretability of transformer models using sophisticated visualization techniques.

Software Engineer

• May 2022 - Aug 2022

- Wrote data pipelines to easily facilitate training of machine learning models. - Implemented a tool to detect, track and correct the behavior of malicious data annotators in Instagram Shopping in Hack. - Trained and deployed machine learning models to reduce inappropriate content on Instagram Shopping. - Wrote data pipelines to easily facilitate training of machine learning models. - Implemented a tool to detect, track and correct the behavior of malicious data annotators in Instagram Shopping in Hack. - Trained and deployed machine learning models to reduce inappropriate content on Instagram Shopping.

Software Engineer

• Aug 2020 - Nov 2020

- Engineered an end to end system to train and deploy object detection models. Single handedly developed the model training framework using REST APIs which supported 3 object detection frameworks. - Completely automated the process of model training and deployment across 3 different platforms (GCP, AWS and local) by creating a software package used internally by the data science team as well as by clients. - Performed pruning and quantization and deployed deep models on mobile and edge devices using TFLite and TensorRT. - Programmed an application to convert images and annotations to one of 5 different formats. - Increased mAP by 5% while getting 2x the fps from the object detection framework. Show less

Software Engineer

• Jun 2019 - Mar 2020

- Led a team which designed and implemented a Content Based Image Retrieval system which fully automated the retrieval of similar images from a data set of 400,000 images. Reduced the time required by a team of human workers to find similar images from 1 week to an automated system which retrieves images in a few minutes. - Developed a system to perform aspect based sentiment analysis in hindi on ecommerce reviews. - Led a team which designed and implemented a Content Based Image Retrieval system which fully automated the retrieval of similar images from a data set of 400,000 images. Reduced the time required by a team of human workers to find similar images from 1 week to an automated system which retrieves images in a few minutes. - Developed a system to perform aspect based sentiment analysis in hindi on ecommerce reviews.