Abhishek Prakash

Building AI chips at AWS

Seattle, WA --:--

All work

Digital Pricing Research Platform

Amazon · Software Development Engineer II

Highlight
p99 latency under 100 ms
Company
Amazon
ML pricing and experimentationThe transition from static pricing rules to machine learning and A/B experimentation, shown conceptually. These shapes are illustrative and do not depict a proprietary hardware design or production topology.ABStatic rulesML pricingA/B experiments

Overview

I led the re-architecture of a static, rule-based pricing engine into an ML-based platform with A/B experimentation. I partnered with scientists on platform design and owned the experimentation architecture, while also building customer-facing pricing services and improving the underlying AWS infrastructure.

Key contributions

  • ML pricing and experimentation architecture: Designed the platform architecture with science partners and owned the A/B experimentation architecture used to evaluate ML-driven pricing strategies.
  • Reference Pricing backend: Designed and built a high-performance backend that enabled Amazon’s digital businesses to derive and display reference prices on product detail and search pages, maintaining p99 latency under 100 ms.
  • Pricing-service performance and cost: Implemented architectural and AWS infrastructure improvements that reduced pricing-service latency and delivered $1.2M in annual operating cost savings.

Skills

  • ML platforms
  • A/B experimentation
  • Distributed systems
  • Low-latency services
  • High-throughput services
  • Performance optimization
  • AWS infrastructure

More work

All work