Case Study: PipeSearch achieves a searchable global tubular inventory through ML-driven data modernization with Rackspace Technology

PipeSearch transforms the global piping and tubular trade

PipeSearch is a U.S.-based software company that connects global steel pipe suppliers with demand, offering intelligence, analytics and QA/QC services. It faced a major industry challenge: demand fulfillment relied on inconsistent, unstructured inventory descriptions (spreadsheets, PDFs, CSVs) with no standard identifiers, making it hard to match technical requirements to available stock across hundreds of suppliers.

Partnering with Onica (a Rackspace Technology company), PipeSearch built a data pipeline on AWS—using S3, Lambda and Amazon SageMaker—and trained a blank NER-style machine learning model on millions of records to normalize inventory attributes. The model reached over 80% confidence in week one and improved with iterative retraining, producing a searchable global inventory that expanded PipeSearch’s platform capabilities, reduced lead times and costs for buyers, and created a digital sales channel for suppliers.

PipeSearch

Briggs Thompson

Co-Founder