Fayzulla Abdurakhimov

About Fayzulla Abdurakhimov

My interest in data and decision-making began early in my academic journey, eventually leading me to pursue a Master’s degree in Business Analytics & Data Science at the University of Texas at Dallas. I’ve always been drawn to understanding how complex systems behave, and how data can shape smarter, more scalable solutions.I joined Amazon in 2024 as a Business Intelligence Analyst within the FBM team, where—starting as a team of eight—we delivered high-impact projects that improved seller efficiency, on-time delivery enforcement, and multi-location fulfillment. I worked extensively across AWS services (EC2, S3, Lambda, SageMaker), Redshift, Spark, and Quicksight to build data pipelines, dashboards, and operational insights used across the organization.In June 2025, I had the privilege of joining Project Amelia, Amazon’s GenAI initiative, where I currently focus on developing analytics, experimentation frameworks, and performance insights for our AI agents (TextAPI, Text2RAG, Text2GraphQL). Our work helps millions of sellers automate tasks, improve decision-making, and adopt next-generation AI tools confidently.Originally from Uzbekistan and now based in Seattle, I enjoy outdoor activities, fitness, and continuously challenging myself to grow as a data professional and AI builder.I am fully open for new connections. Please feel free to reach out if you need help or advice!

Experience

1. Business Intelligence Analyst

   Amazon

   Jun 2024 — Present · Seattle, WA, US

   Selling Partner Experience (SPE): Project Amelia - Amazon\'s AI assistant for sellers-Automated GenAI management reports by optimizing ETL pipelines (SQL Redshift, Datanet) and data models, reducing Excel load by 400K rows, enabling real‑time product metrics (engagement, adoption), and saving 8 hours per week-Improved GenAI assistant engagement through A/B testing using Python and experimentation framework across 3 query‑routing strategies, reducing seller contacts by 10% and boosting adoption by 15% and reaching 1.2 million active users-Optimized ETL performance by automating the top 3 longest‑running jobs using partitioning, sort keys, and filters (rundate, region_id) in Spark cutting runtime by 140 min/week per job and enabling faster access to decision‑critical data-Delivered unified data model and dashboard which track latency, benchmarking, engagement of different AI agents (8 tables, 65 enforcement metrics) in Quicksight, enabling leadership to act on 20K low‑performing sellers weekly and improve compliance metrics-Maintained topic modeling dashboard (Quicksight) tracking GenAI assistant interactions and partnered with Product, Science, and Engineering teams to identify top‑performing agents, increasing relevant responses by 12% and reducing low‑quality interactions by 8%.

Education

• The University of Texas at Dallas

  Master of Science in Business Analytics, Data Science

• Inha University in Tashkent

  Bachelor of Science in Business Administration in Logistics