Apr 2023 – Aug 2024 · 1 yr 4 mos · Bengaluru, Karnataka, India

Mar 2022 – Apr 2023 · 1 yr 1 mo · Bengaluru, Karnataka, India

Apr 2021 – Feb 2022 · 10 mos · Bengaluru, Karnataka, India

Jan 2021 – Mar 2021 · 2 mos

Jul 2020 – Jan 2021 · 6 mos

May 2020 – Jul 2020 · 2 mos

Jun 2018 – Feb 2020 · 1 yr 8 mos · Bengaluru Area, India

• Adobe Media Optimizer Team-
• Adobe Media Optimizer(AMO) is an ad buying and management platform which manages ads across various search and social engines.
• Social Advertising Project - Extended the backend framework to support data synchronisation and updates between Facebook Marketing API and AMO.
• Search Advertising Project - Working on RESTApi frameworks and also handling the ad tracking APIs between Search Engines and AMO. Also, exploring how an efficient data processing pipeline can be built for AMO mainly using frameworks like Airflow.
• Data Analysis Platform (Frontier) - Working on python and R scripts which is used to run simulations
• on how the AMO platform will be beneficial to the client’s ad systems by providing them information on cost, revenue etc.
• Technical Stack: Python, Flask, PostgresSQL

Jan 2018 – Apr 2018 · 3 mos · Bengaluru Area, India

• Worked on a project which involved Enhancing Images using Convolutional Neural Networks. This project was based on two contexts, one being removing artefacts due to the acquisition of an image and the other being improving image quality to make it more visually pleasing. Worked on two problem statements -
• > Joint HDR and Depth Estimation from varying exposure stereo image pairs
• > Enhancing Images to make the objects more recognizable by state of the art CNN's.
• The aim of the first projects was to find a effecient Deep learning based method to generate HDR image as well as corresponding accurate depth maps from varying exposure stereo pairs taking into consideration the ghosting artifacts. I explored many CNN's for this method like Monodepth,Siamese Networks etc. Extensive experiments were done on these CNN's were done.
• The second project focused on finding a more recongnizable image from a raw image. I explored the technique of GAN's to generate enhanced or transformed images given a raw image. Experiments were done by training the generator in a network containing pretrained and finetuned state of the art models like VGG's, Resnet etc.

May 2017 – Jul 2017 · 2 mos · Bengaluru Area, India

• Worked at Adobe BEL Labs under the supervision of Dr. Iftikhar A. Burhanuddin in a team of 4 members on research projects based on Machine learning for innovating the products of Adobe. The project was titled Privacy Aware Recommender Systems. The problem statements were -
• Incorporating personalization in existing differentially private recommender systems
• Differential privacy in sequential based recommender systems
• These problem statements required extensive mathematical approaches to modify the internal procedures of the methods used. Mainly, the work was based on using LSTM in recommender systems by modifying the LSTM optimization to incorporate privacy. It also involved implementing these architectures from scratch. Also, some reinforcement learning concepts like differentially private policy gradient, Q-learning were also explored.
• Patent (Application No. 16041182) titled 'Recommendation System Based on Individualized Privacy Settings' issued at the US Patent office.

May 2016 – Jul 2016 · 2 mos · Kolkata Area, India

• Worked on Robust Minimum Divergence Estimators indexed by one or two tuning parameters. Explored the efficiency and robustness of the estimators through extensive simulations and real data Analysis on R. These estimators may be computed without any non-parametric smoothing of data

Jun 2015 – Jul 2015 · 1 mo · Jaipur Area, India

• Worked on projects based on Web and App development. Developed Web and Mobile based apps with the help of platforms like PHP, Parse and Android Studio.