Amazon

Software Engineer

reducing manual processing time from 2 days to instantaneous

an automated file validation microservice for custom exports utilizing .NET framework

fostering seamless interaction among providers and patients

Designed and implemented an automated file validation microservice for custom exports utilizing .NET framework, reducing manual processing time from 2 days to instantaneous, fostering seamless interaction among providers and patients

through extensive testing and feedback loops

ensuring error-free releases in an agile environment

Reduced rollbacks by 40% through extensive testing and feedback loops, ensuring error-free releases in an agile environment

build custom agents trained on GloVe Embeddings and LSTM models

achieving a 90% response accuracy in providing medication guidelines

Utilized LLM framework LangChain to build custom agents trained on GloVe Embeddings and LSTM models to translate natural language into executable backend achieving a 90% response accuracy in providing medication guidelines

Implemented Global Secondary Indexes on DynamoDB by analyzing backend query routes reducing query times by 40%

using Python and JavaScript, based on a potential profit algorithm

expediting lead generation for the Regulatory Sciences team

Developed a web-based prospecting tool using Python and JavaScript, ranking 50+ biopharmaceutical companies based on a potential profit algorithm, expediting lead generation for the Regulatory Sciences team

a large dataset of ingredient images, reaching an accuracy of 94.29% on the test set

using TensorFlow, trained on a large dataset of ingredient images

reaching an accuracy of 94.29% on the test set

Engineered a custom image recognition model using TensorFlow, trained on a large dataset of ingredient images, reaching an accuracy of 94.29% on the test set

using U-NET and FRCNN, in collaboration with a dental resident

published findings in International Journal of Computerized Dentistry, 2023

Orchestrated development of a 90% accurate teeth segmentation and labeling software for oral scans using U-NET and FRCNN, in collaboration with a dental resident, and published findings in International Journal of Computerized Dentistry, 2023 (View Publication).

95% Dice Score, Silhouette Coefficient of 0.5

instance segmentation on CARLA road image dataset with U-NET, K-Means and Gaussian Mixture Models

enabling more accurate perception for autonomous vehicle navigation

Performed instance segmentation on CARLA road image dataset, securing a 95% Dice Score with U-NET and a Silhouette Coefficient of 0.5 with K-Means and Gaussian Mixture Models, enabling more accurate perception for autonomous vehicle navigation.

using Full Stack, MongoDB with AWS, engaging with multiple customers, designed process matrix, incorporating incident management system

Developed an ecommerce platform using Full Stack, MongoDB with AWS by engaging with multiple customers and designed process matrix to scale up the business. Optimized with best measures with a page TBT response time less than 70ms and enhanced the productivity by incorporating incident management system with priority (P1, P2, P3).

increase in user engagement and a significant drop in bounce rates

comprehensive website redesign, focusing on intuitive navigation and user-centric design

notable increase in user engagement and a significant drop in bounce rates

Spearheaded a comprehensive website redesign, focusing on intuitive navigation and user-centric design, leading to a notable increase in user engagement and a significant drop in bounce rates.

Revamped the employee web portal and iPad application with React, JavaScript, CSS, and HTML, achieving a 50% reduction in load times and improvements in user satisfaction.

web scraping application with the Flask framework

Developed an efficient web scraping application with the Flask framework to fetch and update live prayer times from the University of Southampton ISOC website used by 500+ society members.

using Flask, which facilitated web scraping to deliver real-time prayer times, integrating stream processing for efficient data management

Developed RESTful API endpoints using Flask, which facilitated web scraping to deliver real-time prayer times, integrating stream processing for efficient data management, ultimately achieving an API response time of 150ms.

implementing the generation and propagation of security tokens

Strengthened security measures in a team API by implementing the generation and propagation of security tokens, increasing protection for customer data across ~120 transactions per minute

introducing a UI development tool to a non-servered component package

Achieved a ~9 minute reduction in manual testing overhead per code-change by introducing a UI development tool to a non-servered component package, increasing developer efficiency

20+ responsive web pages, enhancing user engagement by 30%, reducing page load times by 25%

using HTML5, CSS3, JavaScript, jQuery, Angular, and ReactJS

enhancing user engagement by 30% and reducing page load times by 25%

Developed 20+ responsive web pages using HTML5, CSS3, JavaScript, jQuery, Angular, and ReactJS, enhancing user engagement by 30% and reducing page load times by 25%.

achieving a 30% increase in transaction speed

Created an e-commerce site using JSP, Servlets, and EJBs, achieving a 30% increase in transaction speed.

to Java based microservices deployed on AWS EC2 using Docker

Migrated monolithic applications to Java based microservices deployed on AWS EC2 using Docker, achieving a 40% reduction in deployment time.

streamlining access to insurance contract management for buyers

Developed cross-platform customer portal using React, Flutter, and Node, streamlining access to insurance contract management for buyers and achieving a 40% efficiency increase in contract handling

analytics dashboard, predictive modeling, PostgreSQL data indexing

Created analytics dashboard with insights into contract spending, predictive modeling for future contract pricing, and price fluctuations, achieving a 35% reduction in data retrieval times through PostgreSQL data indexing