Nonprofit / Procurement

From Manual Workflows to Intelligent Product Matching

Intelligent Product Matching replaces hand-checking supplier sites, manual match qualification, and spreadsheet record-keeping with automated catalog search, AI help on close calls, and one place to decide with proof.

Year 2026

See the story Start a similar project

Supplier
Websites

Spreadsheets
& Records

Manual
Checking

Recommendation

Match

Automated search

AI on Close Calls

One Place to Decide with Proof

+70%

Less time spent hunting look-alikes across supplier sites by hand

+60%

Improvement in match qualification speed

+75%

Less spreadsheet re-entry to keep match records

+50%

Less manpower burned on the find–qualify–record loop

Overview

The problem, the approach, the result

Challenge

The matching work was almost entirely manual: staff checked look-alike products on other supplier sites one by one, judged by hand whether each item qualified, then typed the results into spreadsheets. That loop burned time and headcount — and still depended on who remembered which tab. Supplier sites also each work differently, so one-size-fits-all hunting kept missing products or pushing people straight back to copy-paste.

Solution

We built an automated matching workspace. Teams load their product list, choose supplier catalogs, and get ranked look-alikes without opening every site by hand. Strong matches rise for review. Close-but-unclear matches can get an AI second opinion. Reviewers compare side by side, keep notes and proof in one place, and hand off the decision — so the sheet is a starting list, not the forever home for every match. If automation misses a corner of a site, staff can still save the item from their browser into the same workspace.

Outcome

Find, qualify, and record stop being three manual jobs. People spend time judging real candidates — not hunting tabs or re-typing rows — and every decision can travel with notes and proof attached.

Automate

Stop opening every supplier site by hand.

Some catalogs are simple lists. Others only show products once the live site is open. Some need a direct supplier connection. Big retail sites hide items across huge store maps. It searches eight catalogs automatically — each the way it actually works — so the first pass no longer burns people hours.

1Search eight supplier catalogs together automatically
2Read each catalog the way it actually works
3Try again another way when the first pass misses items

CABVI automated supplier catalog search screenshot for product matching software

Qualify & record

AI helps qualify close calls — the record stays in one place.

Strong matches rise first so people are not hand-qualifying noise. When something looks close but unclear, AI can offer a second opinion. Reviewers compare side by side, keep notes and captures, and move the decision forward without rebuilding the story in a spreadsheet.

1Rank look-alikes so strong matches rise first
2Use AI when a match is close but unclear
3Compare, keep proof, and hand off without a new sheet