Infrastructure Market Rate
Eliminating the need for manual mappings on the Service Cost Tool
Background
I'm the lead and sole designer on 2 product spaces in the actuals and automation space at AWS Fintech, where I own the full-cycle design process — from strategic roadmap planning to research, design, and stakeholder management — for the Service Cost Tool (SCT).
In 2024, AWS Fintech launched 3 enhancements to Infrastructure Market Rate (IMR) capabilities. I designed the enhancement to the setting process that introduces self-service mapping of a marginal cost product with a public price product.
Impact
The enhancement significantly reduces the need for manual mappings during each rate-setting cycle — manual mapping previously accounted for over 90% of billing errors in 2023.
UX process timeline
Context
The AWS Service Cost Tool is the single source of truth for service unit costs, providing an interactive and transparent environment for users to manage and explore inter-service cost allocations. The enhancement I designed enables users to map product offerings defined within the tool directly in the interface for downstream billing and usage systems.
Challenge
IMR Finance users previously had to manually maintain mapping in an Excel file with ~4MM data attributes and extract data across 100K+ product/region intersections between AWS Fintech systems to support correct billing and analysis. Teams spent ~1,800 hours/year during the rate-setting cycle, and any missing space, misspelling, or formatting issue would stop rates from flowing downstream correctly. There was also no review/approval workflow for IMR overrides.
Solutions
I worked closely with developers and customers across 4 UX workshop sessions to understand pain points and requirements, iteratively created 47 screens to ship this feature, and tested them in the workshop sessions. I built the workflow from scratch in Figma, providing a user-defined mapping process with dropdowns and the ability to map to billing data attributes — dramatically preventing manual errors and removing the need to consolidate and review data in Excel.
Due to the internal nature of this team, I'm limited in what I can share of the actual designs here. Please reach out directly to learn more!
Personal takeaways
This project taught me the importance of early communication with developers, given the technical limitations on the product's backend capabilities that affected the possible user experience. Since this project was an intake request, I learned to stay close to our customers through sessions to ensure every requirement was gathered, and that the solution would scale to other groups of customers.
This was the first project I worked on where I got to see the full launch — which taught me to uphold UX standards even after handoff, ensuring the proper experience is represented at every stage from handoff to UAT to final launch.
Thank you to Elizabeth, Mengze, Jolie, and Hao for the smooth collaboration on this project, from start to finish!