Data Configuration
This is where your portal's product knowledge comes from. You connect DistiForge to your existing product content (websites, uploaded files, or document management systems) and it automatically extracts and organizes the product information for your portal.
1. Data Source Types
DistiForge supports three types of data sources:
Website
Crawls a website to find product pages and PDF documents
File Upload
Lets you upload PDF? docs OR csv files directly
Cognidox
Connects to a Cognidox document management system
Adding a New Data Source
All data sources are created through a step-by-step wizard:
Go to Data Configuration within your portal.
Select the source type: Website, File Upload, or Cognidox.
Configure the source (details for each type below).
Set the default confidentiality level for documents from this source.
Review your settings and click Create Source.
After creation, DistiForge immediately begins processing the source. You can track progress from the Data Configuration page.
A website source crawls a website to discover product documentation and pages.
DistiForge starts from the URL you provide and automatically discovers product content across the site. It prioritizes PDF documents such as datasheets and application notes, and also processes relevant product pages. Content that does not appear to be product-related is automatically filtered out.
Configuring a Website Source
When you create a website source, you will be asked to provide:
The crawler automatically navigates through the site within reasonable limits to cover your product catalog. If your catalog is spread across multiple subdomains or sections, consider creating separate sources for each area.
You can Enable automatic sync, so that your sources are always up-to-date
Password-protected PDFs and pages that require login cannot be accessed by the crawler.
File Upload lets you bring in individual PDF documents directly - useful when your product documentation is not publicly available on a website.
Uploading Files
Create a new data source and select File Upload.
Drag and drop your PDF files into the upload area, or click to browse for files.
Click on Continue
You can review your data and click on Finish
Add Files to the same source
Click on the source, that is already existing (ex. Product A), and then on Add document
Only PDF files are supported. Make sure your files are not password-protected and contain selectable text (not just scanned images).
If your organization uses Cognidox as a document management system, DistiForge can connect directly to it and sync product documentation automatically.
Before You Begin
You will need to set up the Cognidox integration credentials first. Go to Settings > Integrations > Cognidox and enter your Cognidox instance URL and access token. See Integrations for details.
Creating a Cognidox Source
Create a new data source and select Cognidox.
Use the category browser to navigate your Cognidox category tree and select which categories to sync.
Click Create Source.
Automatic Sync
Cognidox sources can stay up to date automatically. When auto-sync is enabled, the tool periodically checks for new or updated documents in the selected categories and brings them in without any manual action needed.
Deleting Documents
To remove specific documents from the knowledge base:
Go to the source, select the documents using the checkboxes in the list.
Click Delete Selected.
Confirm the deletion.
Note that deleting documents does not automatically delete the products that were extracted from them. Those products remain in the catalog and may need to be reviewed and removed separately from the Product Catalog.
Deleting a Data Source
To remove a data source entirely:
Navigate to Data Sources.
Click on the three small buttons an select Delete
Confirm the deletion.
Deleting a source removes the source configuration and all documents associated with it. However, products that were previously extracted from those documents are not automatically deleted, they remain in your product catalog. Review and remove any unwanted products from the Product Catalog after deleting a source.
2. Taxonomy
The taxonomy is the category structure that organizes all products in your portal. It determines how products are grouped for browsing and filtering, and guides the AI when it classifies new products during extraction. A well-organized taxonomy makes your portal easier to navigate and improves the quality of AI-extracted product data.
How the Taxonomy Is Organized
The taxonomy has three levels of structure:
Groups - Top-level containers that hold related categories. These appear as navigation sections in the portal (for example, "Microcontrollers & Processors" or "Power & Analog").
Categories - Product groupings within a group (for example, "MCU," "Power Management," or "Sensors"). Categories can be nested, a category can have subcategories.
Facets - Filterable attributes attached to a category that let users narrow down products within it. For example, a Microcontrollers category might have facets like "Core Type," "Supply Voltage Range," and "Package Type."
Default Categories
DistiForge ships with a set of default categories designed for the semiconductor distribution industry:
Analog
Amplifiers, comparators, and mixed-signal ICs
Connectivity
Wireless and wired communication ICs
Discrete
Transistors, diodes, and MOSFETs
EVK
Evaluation kits and development boards
MCU
Microcontrollers
MPU
Microprocessors
Memory
Flash, SRAM, and EEPROM
Power Management
Regulators, converters, and power management ICs
Sensors
Environmental, motion, and optical sensors
You can use these as-is, customize them to fit your product range, or replace them entirely with your own structure.
Managing Categories
Navigate to Data Configuration > Taxonomy in the admin console to view and edit the full category tree.
Editing a Category
Click any existing category in the taxonomy editor to open it. You can update the category with new one's (button: Add subsystem) or delete them name, parent, group, and facets. Changes are reflected in the portal immediately — including any products currently assigned to that category.
Deleting a Category
Open the category you want to remove.
Click Delete Category.
Confirm.
Deleting a category does not delete its products. Any products that were assigned to the deleted category will have an unassigned category and should be remapped. You can do this from the Product Catalog by filtering for products without a category.
Facets
Facets are the filter options that appear when a user browses products within a category. They let visitors narrow down results by specific attributes, for example, filtering MCUs by core type or voltage range.
Adding Facets to a Category
Open a category in the taxonomy editor.
Click Add Facet.
Choose the Attribute and Values
Click on Add.
Refreshing Filter Options
After publishing a batch of new products, the filter values shown to portal users may be out of date. To refresh them:
Go to Taxonomy.
Click Aggregate from Facets.
This scans all published products in each category and updates the available filter values to reflect your current catalog. Run this whenever you publish a significant batch of new products.
How the Taxonomy Affects the Portal
The taxonomy shapes the portal experience in two ways:
Navigation and filtering: Portal visitors browse products using the category tree and facet filters. Only categories that contain at least one published product are shown. Empty categories are hidden automatically.
AI extraction accuracy: When the AI processes new documents, it uses your taxonomy as a reference to classify products correctly. The more clearly your categories are named and structured, the more accurately products will be assigned during extraction. This reduces the amount of manual category correction you need to do after ingestion.
3. Confidentiality
Here you can define confidentiality levels for your products. These levels control how products are classified and displayed. The AI extraction process will automatically assign a confidentiality level to each product based on these definitions.
You can choose between predefined Levels
Public: Information and sources are freely available with no access restrictions. Anyone may view this content.
NDA: Information and sources are strictly confidential. Access is limited and is requires NDA.
Internal: Information and sources are restricted to company personnel only. This level encompasses all internal access tiers.
You can choose the icon, color, label and more by simply clicking on the level you want to modify. Don't forget to click on save.
How to add a level of confidentiality
Go to Data Configuration > Confidentiality and select Add Level (on the upper right). Fill in the Label, Key, Description (optional) and choose the icon and the color you want.
Delete a Confidentiality Level
Choose the level you want to delete, and click on the bottom left on 'delete', then confirm deletion.
Last updated