Rahul Roy

Built Thai Copilot, an AI-powered Thai language learning application, owning product, backend, frontend, infrastructure, and deployment end-to-end. Developed the application using Ruby on Rails, React, AWS, and Terraform, including interactive language exercises and context-aware learning flows. Consulted with an early-stage AI startup on product feature delivery across backend, frontend, and AI-enabled workflows.

Re-architected B2B partner email notification workflows by decoupling partner configuration from core application logic, reducing new partner onboarding time from weeks to hours. Improved white-label onboarding launch velocity by 5-10x through architectural refactoring, reusable configuration patterns, and cleaner ownership boundaries. Owned end-to-end ADA compliance for B2B post-booking emails, improving accessibility readiness across partner-facing communication flows. Automated B2B email review workflows, reducing manual review effort from 23 hours to minutes and accelerating cross-functional iteration between engineering, product, and operations. Worked across distributed engineering, product, and operations teams to improve reliability and delivery speed in a high-scale travel technology environment.

Owned development and production reliability of Groove's Dialer application as one of two engineers on the squad, supporting a customer-facing real-time communication product. Led implementation of Voicemail Box and International Local Dial features using Ruby on Rails, ReactJS, Twilio APIs, S3, Apex/Salesforce, and ElectronJS. Delivered features that contributed to retention of high-ARR enterprise customers by solving critical calling, voicemail, and international dialing workflows. Designed and maintained GraphQL and REST APIs consumed by frontend and Salesforce integrations, with focus on usability, reliability, and performance. Supported production debugging and incident response for real-time communication systems, converting operational learnings into long-term reliability improvements. Collaborated asynchronously with product, frontend, backend, Salesforce, and security stakeholders across a globally distributed team.

Led architecture and development of VueOps Drive, a high-performance document management system built with Ruby on Rails, ReactJS, PostgreSQL, Redis, and AWS, reducing dependency on third-party vendors. Spearheaded migration of a monolithic Rails-based file processing engine to serverless infrastructure using NodeJS, AWS Lambda, DynamoDB, SNS, API Gateway, and Serverless Framework. Reduced file-processing infrastructure cost by 5x and decreased processing failures by 3x through the serverless migration and reliability-focused redesign. Optimized PostgreSQL queries, Redis caching, and background processing patterns to support high-volume document workloads reliably. Drove technical architecture decisions and engineering standards that improved scalability, reliability, and development practices across the company. Led hiring, onboarding, and mentorship for junior engineers while establishing backend best practices within the team.

Built internal applications for pricing, catalog, content, and marketing teams using Ruby on Rails, MySQL, and AWS, delivering 25x productivity improvements and reducing operational errors by 5x. Created a pricing and catalog automation platform that reduced a 12-person manual operations workflow to 2 people. Partnered directly with business operations teams to convert manual, error-prone processes into reliable internal tools.

Sole backend engineer responsible for Rails APIs for Android and iOS applications; delivered a production payment system under tight business deadlines.

Built full-stack Rails applications for early-stage startups, taking products from concept to MVP in remote, async teams.

Built real-time order management and delivery tracking systems using Rails, Redis, and AngularJS for brands including Burger King and Subway.

Integrated 50+ REST/SOAP APIs into Elasticsearch and contributed to Speedlancer, a microtask platform with payment workflows.