William Lourdhu Raj Rayappan

• Autodesk Learn Platform – System Design & Development
  Learn Platform: Designed, implemented, and developed a streaming-based learning platform using React JS, Node.js, Java Spring, Contentful, DynamoDB, MySQL, Snowflake, AWS Lambda, AWS Batch.
  Personalized Learning Experience: Leveraged GraphDB (Neptune) to deliver tailored learning recommendations. Tech Stack: Node.js, Snowflake, Neo4J, Python, AWS Lambda.
  Learn Live Coaching: Designed and developed Learn Live Coaching for premium customers, enabling interactive learning sessions. Tech Stack: Node.js, MySQL, AWS Lambda.
  Best Practices & Optimization: Established design methodology best practices, improving efficiency, scalability, and consistency across projects.

• Radar – Ticketing & Tracking System Enhancements
  Objective: Enhance Radar, a client-server ticketing & tracking system, by improving event processing, messaging, and access control.
  
  Key Contributions:
  
  Implemented micro-batching using Apache Camel aggregation/split patterns, preventing duplicate notifications via email, APNs, and webhooks.
  Developed common messaging components to integrate applications using Kafka as the underlying data framework.
  Created an activity feed web service for seamless interaction with the Radar client application.
  Enhanced access control (ACL) in Neo4j, enabling fine-grained permissions for problems and components.

• Apache Solr Migration 5.x to 8.x
• Fixing bugs in solr search module/proxy/notifications/data migration/web-services applications
• Added dynamic indexing to Solr for product inventory and pricing details from Kafka based systems which got gets data from different sources.
• Spring endpoints are triggered to start listening to Kafka messages and dump it into mongo DB
• Data sources provide product, sku, price, history of prices etc
• Kafka Consumer for topics related to products, price etc
• The encoded messages are produced and sent to kafka
• Spark component reads the documents from mongodb and indexes in solr

Same-Day Delivery Feature Implementation Objective:

Extend Same-Day Delivery (SDD) availability across the customer journey, from homepage to bag , instead of being limited to checkout.

Implemented Same-Day Delivery features across homepage, search, browse, PDP, and bag .

Enhance search & filtering to allow customers to easily find SDD-eligible products.

• Designed & implemented Same-Day Delivery features across all pages, from homepage to bag .
• Developed core components for Search & Browse (S&B), PDP, and common UI elements .
• Implemented Spring Boot based microservices to support Same-Day Delivery functionality.
• Integrated with platform-level services for seamless SDD handling.

Buy online and Pick up in Store Feature

Key Contributions:

Enhance store pickup functionality across Macy's & Bloomingdale's , improving performance, search, and browse experience for pickup products.

Fixed UI & XAPI production defects to enhance user experience.

PDP Componentization – Macy's

Aligned Product Detail Page (PDP) componentization with Macy's' new architecture using XAPI/Web Generation Layer for scalability and efficiency.

Key Contributions:

Implemented Infinispan caching in XAPI to improve performance.

Optimized XAPI service calls , reducing downstream dependencies.

Led domain-level PDP releases , ensuring smooth production deployment.

Resolved critical production issues ahead of the pre-holiday peak

Search & Browse (SnB) Componentization – Macy's

Aligned Search & Browse (SnB) componentization with Macy's' new XAPI/Web Generation Layer (WGL) to enhance performance, SEO, and multi-device compatibility .

Key Contributions:

Developed XAPI endpoints to serve SnB functionality and dynamic data models.

Implemented server-side rendering for SEO and client-side rendering for interactive UI.

Collaborated with UX architects and stakeholders to define a common data model .

Optimized SnB performance using Apache JMeter , ensuring scalability.

Technologies: XAPI, WGL, Java, Spring, Apache JMeter .

• Header & Footer Componentization – Macys
  Implemented first-of-its-kind componentization using XAPI/Web Generation Layer (Polaris WGL) to create a scalable, maintainable header/footer architecture.
  
  Key Contributions:
  
  Developed XAPI endpoints to provide a context-driven data model for WGL.
  Standardized header/footer, replacing 15 variants for easier maintenance.
  Implemented server-side rendering for SEO and client-side rendering for dynamic UI.
  Collaborated with UX architects & stakeholders to ensure seamless integration.
  Performance tested the HnF stack using Apache JMeter for scalability.
  Technologies: XAPI, Polaris WGL, Java, Spring, Apache JMeter.

Shopping Bag Componentization – Macys
Led the componentization of the Shopping Bag, a key step in transitioning Macys to a microservices-based architecture while ensuring seamless integration and fallback mechanisms.

Key Contributions:

Designed and implemented a standalone Bag service for Macys/Bloomingdale’s with seamless fallback to the full-stack application using F5 iRules & experimentation logic.
Ensured zero disruption to ongoing development by maintaining sync with the existing stack.
Optimized REST services to support the componentized architecture.
Successfully handled 750 TPS during Black Friday with 99% bug-free production deployment.
Coordinated across architecture, infrastructure, support, and release teams to ensure smooth rollout and production stability.
Technologies: Microservices, REST, F5 iRules, Java, Sprin

Dynamic Killswitch System (DKS) – Feature Flag API

Objective: Replaced the existing MBean-based killswitch server with a highly available and centralized feature flag API , enabling dynamic feature toggling in production.

Starting from inception to implementation. This was making use of Zookeeper storage and services are based on Java Spring.

Key Features:

Feature Flag API for client apps ( navapp, shopapp, bagapp ).

High Availability : Ensures functionality even if Zookeeper goes down.

Centralized Control : Features managed across channels, releases, and environments .

Seamless Release Management : Supports smooth application rollouts.

MASS Integration : Macys' monitoring tool (MASS) controls feature enablement.

Note : Similar to LaunchDarkly , but customized for Macys .

• Buy Online, Pick-Up In Store (BOPS) – Macys & Bloomingdale’s
  Led a team of six (including two onsite) to develop a Zookeeper-based solution for managing feature flags and improving BOPS (Buy Online, Pick-Up In Store) functionality, enhancing customer convenience while driving in-store foot traffic.
  
  Key Contributions:
  
  Designed & implemented a POC for a Zookeeper-based solution before project inception.
  Developed XML structures to define feature configurations across channels, releases, and environments.
  Created architectural, design, and support documents for seamless implementation.
  Developed Apache Curator-based clients to manage BOPS-related services.
  Enabled IP-based store selection, allowing first-time users to set a pickup store dynamically.
  Implemented store availability & cutoff time visibility for BOPS-enabled products.
  Enhanced user experience by enabling store selection based on City, State, Zip Code, and Radius.
  Integrated Ship/Pickup toggle options in the Shopping Bag Page for a seamless checkout experience.
  Technologies: Zookeeper, Apache Curator, Java, Spring, XML.

Solution & Implementation: Roles & Responsibilities:

Optimize macys.com and bloomingdales.com for search engines by implementing a tag cloud on product and category pages. This enabled search engines to index static keyword links , improving rankings and increasing organic traffic.

Proposed Spring Integration for seamless interaction between subsystems.

Designed and developed an End-to-End solution ensuring high availability & scalability for processing 200-400M keywords across products and categories.

Automated workflows for long-running asset generation jobs , publishing to staging & production.

Developed Maven-driven deployable artifacts for different environments.

Part of a 10-member onsite-offshore team handling architecture, development, testing & release.

Led performance tuning, production troubleshooting, bug triage, and user story analysis.

Ensured scalability and automation of SEO workflows, reducing manual intervention.

• This project will change the customer experience when navigating/browsing Macys.com as well as when performing product searches
• Faceted Navigation will allow the customer to dictate exactly the types of products she would like to see and drill down through varying attributes based on personal preference
• The implementation of Faceted Navigation and Faceted Search is dependent on the implementation of RDPP, ADN (Mapping Service, New Data Product Categories) and Stella
• BDB Refactoring and Performance Testing
• Updated the complete unit test cases for BDB

Macy's Mobile Web Development
Developed the mobile version of macys.com, ensuring dynamic UI rendering based on device detection using Spring Themes & WURFL Mobile Detection Framework. The project followed Agile methodology (Mingle Tool) and integrated with Macy's Service Gateway for data retrieval in JSON format.

Key Contributions:

Architected & developed a responsive mobile web application with faceted search-to-checkout functionality.
Integrated SEO link module and finalized architecture with the onsite team.
Created architecture & design documents and set up the development infrastructure.
Developed all controllers to interface with Macy's Platform Services.
Optimized performance of the link module by analyzing thread pool counts and heap settings.
Developed a Jetty testing plugin for Eclipse to streamline development.
Maintained zero-warning policy with necessary build plugins for high-quality code.
Technologies: Spring Mobile, WURFL, Macy's Platform Services, jQuery, RedHat Linux, WebSphere App Server 6.

• Designed the complete architecture for the mobile web application.
• Evaluated mobile frameworks like WURFL for device detection and optimized mobile rendering.
• Proposed & implemented Spring MVC with WURFL for faster and efficient development.
• Enhanced functionality by integrating JavaScript frameworks , extending beyond basic in-page JavaScript.
• Extended support for Z39 Protocol Interface in the Omni Platform (Cengage Learning) .

Omni is a layered content indexing & retrieval platform using the Lucene Framework , operating across PST (Presentation Services Tier), BST (Basic Services Tier), and LST (Lucene Services Tier) in ICE environments for Cengage.

Integrated Z39 protocol for content retrieval in the library environment using Jafer (rewritten for Omni).

Enhanced Jafer with support for Marc Record formats and Omni service calls.

Customized Z39 to support SPOL, GVRL, and Lit Products .

Evaluated Jafer/JZKit for Omni integration.

Spring 2.0, Hibernate, Jafer, Lucene, XML, RedHat Linux, Tomcat .