TEN Enigma @ Amazon
November 2021 - nowRole: Senior Software Development Engineer
Search Title Generation System (Feb 2025 – Present)
Leading the design and production implementation of a system to decouple customer-facing product titles from backend search data at Amazon's scale (4B products targeted for 2026, 16B products targeted for 2027). Supporting the science team in developing the algorithm for generating search title replacements using GenAI by discussing required tooling for experimentation. The system is currently scaling at ~30M products/day (limitation: GenAI capacity and not the scalability of our system; target: 80M / day). Drove alignment with Principal Engineers across Catalog and Search organizations on scoping strategies to reduce generation needs by ~3x and storage architecture decisions balancing engineering needs against infrastructure cost (potential 90% cost reduction for storage).
SEO Product Content Augmentation Tool
Built an LLM-based system to identify product niches and augment content, reducing competition and improving discoverability. Scaled to 1M augmentations/day through 18x throughput gains via parallelization and cost-optimized pre-filters. Drove a 34K-product experiment yielding 11% OPS uplift (~$625K), influencing leadership to refocus the program on high-demand products - resulting in 9M+ augmented titles with an average 56% search impression increase. Project is actively developed - expansion into new groups of products
Search Insights Pipeline (Enigma)
Architected a system extracting search insights into automated actions - creating new products, detecting defects preventing discoverability. Scaled to 50M+ queries/week across 100M+ products, generating $300M+ in incremental monthly revenue in 2025. Designed for extensibility, enabling the Business Team to expand integration points across Amazon. Project is in KTLO