Jayvardhan Singh

About Jayvardhan Singh

At Amazon, I work on a foundational service responsible for filtering content for personalization workflows and widgets across the Amazon storefront handling up to 8 million customer requests per second with sub-20-millisecond latency.I\'m interested in roles where AI meets systems engineering—whether designing infrastructure for ML workloads and scaling AI teams, designing AI powered applications, or architecting backend systems at scale.

Experience

1. Software Dev Engineer

   Amazon

   Dec 2024 — Present · Irvine, CA, US

   Architected and deployed a production GenAI ML Tools platform (AWS Bedrock, RAG) serving internal customers, reducing filtering result debugging time by 30%- Optimized LLM inference serving cost/latency by 15%, achieving a 45% cache hit rate by designing a Redis-backed caching architecture with prefix-aware routing- Engineered scalable batch/streaming ETL pipelines processing tens of millions of records to power LLM retrieval, ensuring data quality and maintaining strict end-to-end SLAs- Developed a low-latency ML model serving system (sub-20ms) for real-time scoring, implementing A/B experiment hooks that improved widget CTR by 20%- Led global infrastructure scaling and resilience initiatives, reducing peak p99 latency and ensuring N+1 region failover for critical storefront traffic.

Education

• NYU Tandon School of Engineering

  Master of Science - MS, Computer Engineering

• Manipal University Jaipur

  BTech - Bachelor of Technology, Computer and Communication Engineering