Turn product catalogs and brochures into structured product, purchasing, and packaging records for your database, PIM, or ERP.
A product brochure is made for people to read. It may contain dozens or hundreds of products, with names, product codes, prices, and other details spread across pages and tables.
But a database needs those details organized into separate fields. A PDF cannot simply be loaded as a collection of product records. Someone usually has to read the brochure and enter each product by hand.
Manual entry takes time, especially when a catalog contains many products. It is easy to miss a detail, copy a value into the wrong field, or enter the same product twice. When a new brochure arrives, the work starts again.
The aim is to reduce that repetitive work so people can spend their time checking product information instead of typing it.
The system reads a product catalog or brochure PDF, extracts facts supported by the document, and organizes them into named fields. This is product data onboarding: preparing information from a document for use in your business systems.
The information falls into three groups:
| Data group | What it means |
|---|---|
| Product Master Data | Product identity and attributes, such as product code, name, and specifications. |
| Commercial / Purchasing Data | Quoted unit price, currency, minimum order quantity, and purchasing terms. These can vary by supplier or offer. |
| Logistics / Packaging Data | Packaging dimensions, cubic volume, and shipping quantities supported by the document. |
This evidence-grounded extraction keeps source-page and source-text references so records can be checked. Missing information should stay empty instead of being guessed; source units and currency are preserved. Once the records have been reviewed and mapped to your database's fields, your integration can save them without manually re-entering each product.
flowchart TD
Input["Product catalog or brochure PDF"] --> Read["Read the document"]
Read --> Extract["Find product details"]
Extract --> Structure["Organize product, purchasing, and packaging data"]
Structure --> Review["Check and review the records"]
Review --> Connect["Your integration saves approved records"]
Connect --> Database[("Your database or product system")]
For example, a brochure entry such as “Oak Chair · CH-101 · Quoted unit price: USD 85” could become:
| Product code | Product name | Quoted unit price | Currency |
|---|---|---|---|
| CH-101 | Oak Chair | 85 | USD |
The code and name identify the product; the quoted price and currency belong to its commercial offer. This is an illustration of the idea; the output guide explains the actual record structure.
The project implements document extraction, structured product and commercial extraction, and some data-quality checks. It returns review statuses for your application to handle.
Extraction recovers facts from the document. Enrichment adds descriptive or merchandising metadata using grounded product information and images. Image-based enrichment currently uses demonstration data; it does not supply missing prices or purchasing terms.
Saving records to your database, PIM, CRM, or ERP requires a custom integration. Your application handles field mapping, human review, approval, and synchronization. See the guides for implementation details and current limitations.
- Product Master Data: the product's identifying, descriptive, and technical attributes.
- Supplier Item Master Data: the supplier-specific representation of an item, with links to its purchasing and packaging information.
- PPV: a project-internal label used in extraction code and fields such as
ppv_results. It covers product identity, commercial offers, and packaging; it is not a synonym for Product Master Data alone. - PIM / ERP: Product Information Management / Enterprise Resource Planning—systems that manage product information and business operations.
The glossary explains these terms, MOQ, CBM, source evidence, and the technologies used in the project.
- Get started: install the project and process your first PDF.
- Understand the output: see what product records and document references contain.
- Read the glossary: business terminology and internal code labels explained.
- Connect your database, PIM, or CRM: map, review, and save the extracted information using your own integration.
- Explore the technical architecture: detailed diagrams, processing steps, and current limitations.