Instagram Relevance Labs.

Drove global expansion of trading platforms by extending to fixed income products and EMEA/APAC regions, and modernized core risk/quoting systems through streaming architecture, enabling faster product delivery and improved cross-region trading from third-party clients.

Position: SDE-3

Kindle Manga – Web Reading

Designed integration strategy and technical architecture for merging Manga Web Reader with Kindle Web Reader to unify development and maintenance for diverse content formats (re-flowable and fixed-format) in alignment with the Reader org.

Championed migration from the web reader's outdated framework and re-architected it for compatibility with core AWS, aligning with company-wide directives. Addressed longstanding tech debts & streamlined the stack for enhanced efficiency.

Selected as an engaged member of the org-wide "samurai" group, contributing to cross-team design oversight. Providing external insights on scalability and alignment with org objectives. Leading design discussions for upcoming reader features.

Position: SDE-2

Kindle Manga – Discovery & Web Reading

Architected an ad-supported reading program for Manga Web Reader using Amazon’s DSP Ad platform. Overcame system incompatibilities through deep-dives and knowledge-shares. Led development of vital features: optimised bid placements, latency-reducing pre-bids, and fraud detection to enhance publisher credibility. Shared expertise with cross-functional teams and contributed to the triumph of the program, serving as a foundation for subsequent ad-backed content initiatives.

Revamped the Web Library to prioritise series-based organisation, enhancing navigation for large libraries (1000+ items). Streamlined the user interface, simplifying content discovery and facilitating up-selling of missing series items. Addressed longstanding user challenges, garnering positive public feedback on social platforms.

Streamlined tech-ops by clearing backlogs, mitigating tech-debt patterns during on-call cycles, and improving test automation for full CI/CD integration. As a mentor, I provided guidance to team members and interns, conducting code/design reviews and fostering feedback loops that led to promotions and full-time positions. Through knowledge-sharing, I expedited onboarding processes, aiding new members in swiftly embracing on-call responsibilities including monitoring & bug-triaging.

Revamped and built an efficient up-selling UI component for the web reader, offering browser performance, innovation flexibility, and seamless integration with the reader. Enhanced service layer for improved data alignment with core reader functionality. Additionally, I elevated Manga page features through A/B tests, optimising content discovery, integrating social sharing, and retiring obsolete platform integrations.

Position: SDE-2

Improving on consumer experience through multiple projects.

GameX

Innovated and crafted novel game mechanics such as "predict and win," pivoting the winner selection model for contests where results were revealed after completion. This was immediately used for that year’s IPL season. Additionally, engineered a real-time quiz structure supporting up to 1 million concurrent players, featuring seamless video stream synchronisation.

Optimised a marketing campaign management platform by streamlining the tech stack and creating a unified data model. By introducing draft and versioning capabilities for campaigns, I was able to eliminate a redundant micro-service. This transformation reduced developer workload by about 30%, lowered hardware expenses, and simplified data intricacies across services. The migration was executed seamlessly, avoiding downtime, data loss, or regressions.

Chatbots – Conversational Shopping

Conceived & implemented an Amazon product page chatbot, swiftly providing customers with information from descriptions, reviews, and Q&A sections, bolstered by a machine learning service to enhance purchase confidence. Furthermore, I restructured the ML module for scalable deployment and established it as a template for other ML-driven services. This initiative significantly curtailed operational complexity for similar ML backed service architectures.