A supplier document library that matches files to products
AppSheet Solutions and the client’s in-house team moved a supplier document library from AppSheet to custom software for a European ingredients distributor.
Illustration. Names and rows are sample data; totals are the project’s.
On this page
Key numbers
20,000+documents matched to products
≈ 8,200files renamed
Measured in the client’s test environment.
The problem
Certificates of analysis, specifications and safety data sheets arrive from suppliers in a shared Drive folder, tens of thousands of them. The AppSheet app filed them, but linking each file to the right product, spotting duplicates and naming files consistently still took manual work.
What we did
Moved the library from AppSheet into the client’s custom platform: 23,000+ existing records, in the client’s test environment
Reading for every file: text extraction, OCR for scans, then AI extraction of document type, product names and codes, batch numbers and dates
A rule-based product matcher with regression tests; uncertain matches go to a reviewer instead of being guessed
One naming pattern for the files on Drive, and topic tags for search
Built together with the client’s in-house team.
Results
20,000+ archived documents re-read and matched to products, in the client’s test environment
About 8,200 files on Drive renamed to one consistent pattern, in the same environment
How it helps
Staff open a product and find its documents. Reviewers can focus on uncertain matches.
Someone had to write up tasks by hand after long online meetings, and management had no simple view of how company Google accounts and files were used.