Quick Summary

Brown Strauss Steel, a structural steel distributor with nine U.S. facilities, replaced hand conversion of customer order PDFs with an AI extraction workflow that loads orders straight into its SyteLine ERP. An upload that took about an hour now takes under ten minutes, across 50,000+ documents a year.

The Use Case

Customer orders arrived as PDFs. Sales reps converted each one to Excel, reformatted the columns to match what SyteLine expected and uploaded the file. One formatting slip could break the whole upload, and the PDF converter tools they relied on failed often enough to cause outages of a day or two. After testing seven or eight platforms on sample documents, the company chose Nanonets. The workflow reads each order, identifies the table structure, pulls the item, length, quantity and piece mark for every line with a confidence score, reshapes the data to the SyteLine schema and imports it.

Real Example

1. A fabricator emails a 40-line order as a PDF with its own column layout
2. The intake step picks it up and identifies the table and headers
3. Each line's item, length, quantity and piece mark are extracted with a confidence score
4. Two low-confidence lines are shown to the rep for a quick check
5. The data is reformatted to SyteLine's import layout and loaded
6. The rep confirms the order in the ERP about eight minutes after it arrived

Why It Matters

• Time per order upload dropped from about 1 hour to under 10 minutes
• Extraction accuracy above 95% on 50,000+ documents a year
• Company reports 5x return with a 3-month payback, and no more converter outages

Source

https://nanonets.com/customer-success-story/brown-strauss-streamlines-sales-operations