Rahul Ashwin Sheth

Amazon

Software Development Engineer

• Part of the Internal Matching Pricing team, developing infrastructure for generating and publishing competitive prices for Amazon products against third-party sellers, using Java, TypeScript, AWS, Docker, REST APIs, Coral framework, and microservices to maximize Amazon's long-term profitability goals.

Lumen Technologies

Software Developer II

• Contributing to building a Data Marketplace platform using Java, Docker, Kubernetes, Apache Kafka. Implemented a robust data pipeline and utilized AppDynamics for real-time data transmission, resulting in a 20% reduction in server response times.
• Leading a seamless migration of Lumen's Cloud Dashboard data to Azure with 0 downtime, implement IAM for self-managed access, and cut platform costs by 40% by transitioning to Azure Service Fabric.
• Spearheaded migration of data marketplace services to Rancher Kubernetes. Conducted iterative load testing up to 600k requests in 10 seconds using JMeter, optimizing CPU and memory configurations.

Arizona State University

Software Developer

• Ameliorated an internal document management system with Academic Enterprise Communication(AEC) team using Flask, React, and Elasticsearch.
• Deployed a daemon service that automates retrieving 6000+ documents for a DevOps database and created a search API using Elasticsearch Query DSL with text analyzers, enabling fuzzy queries with synonyms, typos, and operators.

Arizona State University

Teaching Assistant

• Aided course instructors with 2 major courses titled LES 305 (Business Law & Ethics) and LES 380 (Consumer Perspective Business Law) in creating course materials, assignments, grading, and proctoring examinations with their large lecture classes of more than 300 students.

Accenture

Associate Software Engineer

• Contributed to the design of a back-end server with Java, Spring Boot, and other cloud technologies for a Message Processing Framework. Optimized performance by 30% through Kafka partitioning for concurrent processing, and refactoring into two microservice components, improving database retrieval efficiency by 20% and expediting message transformation/validation.
• Implemented Spring Boot services utilizing AWS, which reduced over 66% manual efforts across 10+ member QA team.
• Formulated APIs to automate 30% of inventory management workflows using AngularJS, TypeScript, and MongoDB.
• Designed and implemented automated build and deployment data pipelines for 23+ microservices with Jenkins.

Gateway Group of Companies

Project Trainee

• Deployed ReactJS dashboard for layout management allowing seamless configuration of Android mobile layouts for customers.
• Collaborated with developers to optimize database design and query performance, resulting in a 20% decrease in average query response time. Successfully migrated a legacy Oracle database to MongoDB, reducing hardware costs by 50% and increasing query throughput by 30%.

Capitalist Softech Pvt. Ltd.

Software Engineer Intern

• Enhanced e-commerce platform handling 50+ million yearly transactions using Java, Scala, and AWS technologies, improving code reviews and automating workflows with scripts.
• Led intern team to configure ELK stack to aggregate logs from multiple systems and applications, analyzed utilizing Kibana visualization, improving troubleshooting by 25% and efficiency by 15%.

ML Based Stock Price Prediction

TensorFlow, LSTM, Keras, CUDA, GraphViz Mar 2022 – May 2022

• Leveraged Python's yFinance library to collect crucial market data from Yahoo Finance, including moving averages, high, low, and stock volume.
• Built a Long Short-Term Memory (LSTM) network for time series forecasting of closing prices, requiring TensorFlow, Keras, and GPU configurations for execution.
• Prepared and preprocessed testing data for its integration into the LSTM model, enabling accurate stock price predictions.
• Conducted a performance comparison between Linear Regression and LSTM, determining LSTM's superior efficacy with a Root Mean Square Error (RMSE) of 0.43.
• Achieved a high prediction accuracy of 96.6% with the LSTM model, indicating its effectiveness in stock price forecasting.

Horizontal Scaling and Advanced Resource Scaling on AWS

AWS S3 Bucket, AWS EC2, Amazon RDS, Amazon ECS, AWS Lambda, Python. Jan 2023–Mar 2023

• Engineered a VM network and invoked cloud APIs programmatically to provision cloud resources for a varying load.
• Configured and launched an Elastic Load Balancer (ELB) along with an Auto Scaling Group on AWS for backend infrastructure.
• Formulated fault-resistant elasticity policies to uphold QoS of web service.

Python Backend Application for AWS CI/CD

AWS S3 Bucket, CodeDeploy, CodeBuild, Jenkins, Python September 2022–December 2022

• Engineered a Python backend application interfacing with AWS services, enhancing the CI/CD process via a custom Jenkins pipeline.
• Crafted an appspec.yml for AWS CodeDeploy, effectively directing installations from Amazon S3 or GitHub repositories and managing lifecycle hooks..
• Formulated a buildspec.yml for AWS CodeBuild, facilitating streamlined installations and optimized artifact handling, leading to more efficient build times.

Currency conversion Microservice using Docker

Java, Spring Boot, Docker, Kubernetes May 2022–Aug 2022

• Developed and maintained the Currency Transformation Microservice, a Java-based application that converts various world currencies, using Docker for containerization and Maven for automation.
• Utilized Kubernetes ConfigMaps for efficient environment variable management, increasing the application's adaptability and improving configuration management.

News Web Application with React and Node

JavaScript, React, React-Bootstrap, Express, AWS, SCSS Jan 2022–July 2022

• Successfully constructed the frontend using React for page routing and incorporated React-Bootstrap with media queries in SASS, achieving a responsive design that accommodated over 10,000 monthly users.
• Efficiently developed the backend utilizing Express, capable of delivering custom news and search results to approximately 20,000 queries per day, enhancing the user experience through personalization.