catlg Open catlg
Supplier catalogue intelligence

Stop fixing supplier spreadsheets by hand.

Suppliers send product data in every format imaginable. Upload it to catlg and get back one clean file: columns mapped, names standardised, duplicates found, wrong photos flagged, and everything ready for Shopify, WooCommerce or your ERP.

Try it with a catalogueSee how it works
  • Excel, CSV and PDF
  • Learns each supplier
  • You review before export
catlg review table showing products with flags such as bad barcode, price of zero and possible duplicate
How it works

From messy file to clean import in six steps.

Fast rules and image models do the heavy lifting. AI is only used for the few things rules can't read, which keeps it quick and cheap.

  1. Extract

    Finds the real header row, fixes broken characters and encodings, reads tables out of PDFs and pulls out photos embedded in the sheet.

  2. Map columns

    "Art.-Nr.", "Bezeichnung", "VK Preis": recognised in English, German, French, Spanish, Dutch and Italian. Unclear columns are checked against their values. Your mapping is remembered for next time.

  3. Standardise

    Consistent names, units (10mm → 10 mm), prices, barcodes, colours and sizes. Brand pulled from the name when the column is empty.

  4. Match photos

    Matches photos by file name first, then by what the picture shows. Catches the same photo used twice and photos that fit a different product.

  5. Find duplicates

    Same product listed twice, even with different wording. Different sizes, colours or model numbers are kept apart.

  6. Flag and export

    Missing SKUs, invalid barcodes, odd prices, cost above price. Fix them in the table, then download for Shopify, WooCommerce or your ERP.

See it

Built to be reviewed, not trusted blindly.

Column mapping screen with confidence labels
Column mapping you can check in secondsEvery guess shows its confidence and sample values.
Duplicate groups side by side
Duplicates side by sidePick the record to keep; blanks are filled from the others.
Photo review with a flagged wrong photo
Wrong photos get flagged"Looks like: Blue round plate". One click to move it.
Export screen with Shopify download
Import-ready filesShopify, WooCommerce or generic ERP CSV/Excel.

Learns each supplier

Confirm a mapping once. The next file from the same supplier is mapped automatically, with no AI cost.

Variant-aware duplicates

M6x20 and M6x25 bolts, or a red and a blue shirt, are not duplicates. catlg knows the difference.

Photo checks on CPU

Perceptual hashes and a compact CLIP model run on plain CPUs. No GPU, no per-image fees.

AI on a budget

Rules first, AI only for leftovers, with a hard monthly cap per user. At the cap, everything keeps working with rules only.

Edit in place

Click any cell to fix it. Checks update immediately; edited cells are highlighted.

Your data, your control

Delete a catalogue, or your whole account, any time. Spreadsheet formulas in exports are neutralised.

Privacy

Your catalogues stay yours.

Supplier price lists are commercially sensitive. catlg keeps what it needs and nothing more.

Hosted in the EUFiles and data are stored on Google Cloud and MongoDB Atlas in Europe.
Minimal AI sharingOnly column headers with a few sample values, scanned PDF pages, and (if you choose) names to tidy are sent to the AI provider.
No ads, no sellingNo tracking cookies and no analytics. Your data is never sold.
Delete anytimeRemove a catalogue or your whole account yourself.

Read the Privacy Policy and Terms of Service.

FAQ

Questions

Which files can I upload?

Excel (.xlsx and old .xls), CSV/TSV and PDF, up to 30 MB. Photos can be embedded in the sheet, referenced by URL, or added as a .zip afterwards.

What about scanned PDFs?

PDFs with real tables are read directly. Scanned pages are read by a vision model, one page at a time, within your monthly AI budget.

Will it change my data without asking?

It standardises (spacing, units, casing, number formats) and shows you what it flagged. Duplicates are never removed until you confirm, and optional AI rewrites are off by default.

How does it decide two products are duplicates?

Exact SKU, barcode or brand + part number first, then similar names within the same brand. Pairs whose size, colour or model number differ are treated as variants.

Which shop systems are supported?

Shopify product CSV, WooCommerce product CSV (including variable products), and a generic ERP CSV or Excel file with standard column names.

Is it free?

catlg is free during beta, with a small monthly AI allowance per user.

Clean your next supplier file in minutes.

Sign in with Google and drop a catalogue in.

Start free with Google