AKARSHAN JAISWAL

AB Your Orders Desktop:

• Contributed in the implementation of the revamped Your Orders (YO) page for US marketplace on the Amazon Business website which has the second highest traffic on the site with 10 million visits in a year. This led to an enhanced customer experience and a reduction of 100,000 customer service contacts annually, translating to significant cost savings in millions for Amazon.
• Lead the design and development of Order Alerts enabling users to approve or change the payments/payment methods for an order directly from the Your Orders page, streamlining decision-making processes and enhancing user control.
• Formulated the technical strategy for expansion of the Your Orders page across 2 regions and 7 marketplaces which included infrastructure setup, delivery milestones and operational readiness activities. This will help Amazon in further saving approx $22 million annually.

AB Your Orders Mobile:

• Contributed in development of GetDeliveries API to seamlessly deliver display-ready order data, optimising accessibility for Amazon Business users.
• Owned the end-to-end implementation of GetOrderIds API, which ensured that only relevant order IDs were returned based on the customer's permissions and access level.
• Developed High-Level and Low-Level Design for Locker & Campus Orders, seamlessly incorporating these functionalities into the Your Orders mobile page.

Enrich Digital

• Contributed in development of Enrich Digital, an innovative solution aimed at enhancing employee retention and reducing role fatigue through skill development pathways, resulting in a significant decrease in the annual employee turnover rate from 40% to 10%.
• Implemented seamless integration of Routing and Auth services, developed tailored user views, managed infrastructure utilising AWS CDK, executed precise CRUD operations, and engineered modular Java code. These efforts collectively fortified system security, enhanced functionality, ensured scalability, and promoted maintainability.
• Integrated Amazon Federate in Digital Enrich which allows Amazonian's to access internal services using amazon managed credentials.

• Developed MaaS visualizer, an analytical tool vital for assessing congestion and emission data across urban routes and stops, facilitating strategic business decisions
• Did end-to-end software development, including microservice design, REST API implementation using Flask, and deployment on AWS EC2 with NGINX and GUnicorn for efficient API handling
• Through meticulous code optimization, I significantly reduced response time from minutes to milliseconds across multiple microservices.

Heavy Machinery RUL

• I led a project aimed at predicting the remaining useful life of multiple heavy machinery components to mitigate catastrophic failures using machine learning and deep learning models.
• Responsibilities included defining the technical and functional architecture, data pipelining, and refinement. I conducted thorough data analysis and visualisation.
• Furthermore, I performed model training and hyperparameter tuning to identify the optimal machine learning and deep learning models for remaining useful life prediction.
• Technologies utilised encompassed Python, Pandas, scikit-learn, Seaborn, Keras for both machine and deep learning, and Keras Tuner.

Azure Worker Safety

• I developed Azure Worker Safety, a tool designed to detect the presence of safety helmets worn by individuals while working.
• Responsibilities included collecting and preparing image data, training the model for safety helmet detection utilizing Azure Cognitive Services, and exposing the model's API to the frontend for seamless integration.
• Key technologies employed were Azure Cognitive Services for model training and Azure Blob for image data storage.