Joydeep Das Email & Phone Number

Seattle, Wa, Us

- Developing scalable solutions for Amazon Prime Video(PV) Website. Working as a FullStack Software Engineer, from making front-end web widgets to developing backend APIs. Technology stack includes JAVA, React.js, Javascript, Typescript, Node.js in front-end; Functional langauge Scala in backend APIs.- Led projects (Tech lead) in migrating 60+ pages and APIs from Perl/Mason to modern Spring MVC (Java) based solutions. Distributed design for millions of footprints per day. - Designed and implemented several client based storage solutions for Prime Video domain.The project had very wide scale impact (in millions).- Involved in devising strategy for integrating MAWS-NAWS services using various AWSservices. Several projects like PV Trailer Autoplay took advantage of it.- Migrated core prime video applications from in-house resources to native AWS. Design and implementation part included working with Amazon S3, CloudFormation, CDK,Cloudwatch, SNS, Lamdas and DynamoDB.

Gurugram , Haryana, In

Developing vision based On-Device AI solutions for Object Recognition. Deep Neural Networks are devised and deployed for 1000+ object category recognition in Samsung Flagship mobile Devices. Caffe framework is used for training DL models and Python language for validation and testing the models.Worked as Back-end Software Engineer for Samsung’s Bixby Vision app, successfullylaunched in S10, Note 9 and A-series smart phones. Developed several C++ modules fordeployment of AI models on mobile devices.

Uttarakhand, In

Graduated from IIT Roorkee with Major in ECE and Minor in Computer Science with an aggregate of 8.96/10 CGPA.

Seoul, Kr

Single Image Super Resolution using Generative Adversarial Networks: It is the task of upscaling a low resolution image to its high resolution counterpart for a large upscaling factor. Developed an end-to-end Deep neural model for the task using Generative Adversarial Networks.Implemented SRGAN in Tensorflow, with modifications in Loss Function and architecture.The work has been used for commercialization in Samsung Mobile Device, S10.

Gandhinagar, Gujarat, In

Developed a novel algorithm for forgery detection in Median Filtered mages. A statisticalapproach focusing on neighbour pixels and their k th order difference was proposed. Matlabwas used for developing and carrying out the experiments.The algorithm was published in 2017 IEEE International Conference on Identity, Securityand Behavior Analysis (ISBA). https://doi.org/10.1109/ISBA.2017.7947704

Bachelor Of Technology, Electronics And Communication Engineering

Secondary Education, Science

Matriculation, Mathematics