Customer Challenge

The auto manufacturing customer was struggling to modernize legacy systems and manual processes into a strategic and scalable approach. This family-owned business, for 20+ years, had never viewed technology as a core part of the business. The first challenge addressed by BUILDSTR was the fitment matching process that ensured aftermarket parts worked on a specific model of vehicle. This fitment data was incredibly valuable, unique in the industry, and the main reason private equity invested in the business. Before modernization, the matching process
took multiple weeks to complete, often failed in the middle and had to be restarted, and required manual, human intervention several times throughout the process.

Solution

BUILDSTR started by using the matching process as a way to show the customer’s leadership, longtime employees, and new private equity owners that a modern AWS-native approach to data processing and analytics would dramatically improve business results. We started with the end in mind and worked backwards toward a technical architecture that included a number of AWS services and open source tools. For starters, AWS Glue for ingestion, transformation, and cataloging of data; this "verity" piece ensured an unquestionable single source of truth for all data and established a foundation of trust.

From there, data was landed into an unrefined zone in an Amazon S3 data lake, which immediately triggered additional transformations and loaded the transformed data into a refined zone in an Amazon S3 data lake.

The data was heavily operations-focused, relating to all parts of the business from marketing to supply chain, and from SAP, Salesforce.com, Backbase, and other data sources. From the data lake, data was warehoused in Amazon Redshift for the more complex and performant query needs and Amazon Athena for simpler and more ad hoc needs. Data from the entire business was reorganized from top to bottom, eliminating every instance of inconsistency (e.g.,
one part of the business calling inventory "INV" and another calling it "warehouse_inventory").

Data was exposed directly to our customer's customers through many robust APIs and over 70 custom applications. Additionally, data was exposed to all internal parts of the business through internal custom web applications and business intelligence via Tableau Server on EC2.

Results

This effort immediately benefited the matching process by taking the timeline from several weeks and constant failures to less than 30 minutes and zero manual intervention. This process became the cornerstone of the customer's technology estate, proving to the board and the front-line employees that it was possible to dramatically improve the status quo. The matching process also provided critical insights to nearly every custom product and API that delivered value to the business.

This modern approach underpinned the eventual transition to adding direct to consumer e-commerce capabilities, because all hurdles to providing matching information directly
in a customer experience were removed. This cloud-native data foundation has become a major reason the customer and their new private equity ownership were able to execute over 40 acquisitions of competitors and suppliers, effectively going from a small player to easily the largest player in the market.

This was accompanied by a period of 70x growth, from $30M to over $2B in annualized revenue. The long-lasting effects of this initiative and the overall partnership with the customer were not only greater insights for running the business day to day, but also provided the ability to onboard acquisitions in weeks instead of months or years - as all data was properly accounted for in the data foundation and the acquired companies' data was simply mapped to the existing, well-thought out model.