Data Management

Manage the cross-reference database, run ingestion, and monitor data quality.

Normalized Parts

172,925

Equivalence Edges

260,254

Raw Parts Pending

—

Total Lookups

Live Data Ingestion

Scrape real bearing data from distributor sites and import the Timken interchange guide.

Priority 1

nodeshk.com

22,000+ individual product pages via XML sitemaps across SKF, NSK, FAG, INA, NTN, Timken, and more. Each product page = 1 bearing with full specs.

Max products:

Priority 2

central-surplus.com

Highest strategic value for surplus/obsolete parts. Scrapes brand catalog pages for part numbers and availability.

Max brands:

Priority 3

bearingworks.com

Clean dimensional data with load ratings. Single-page scrape of /bearing-sizes/ for validating dimensional accuracy.

Enrichment

Timken Interchange PDF

Cross-reference enrichment layer. Upload the extracted text from the Timken Bearing Cross Reference Guide.

Upload the .txt text extraction of the PDF

API

industrialservos.com

Free cross-reference API with 85K+ parts. Covers filters (hydraulic, oil, fuel, air), bearings, belts, seals. 1,000 free requests/month.

Custom part numbers (optional, comma-separated):

Default: Baldwin, CAT, Donaldson, Fleetguard, WIX, SKF, Gates, John Deere, Komatsu parts

Puppeteer

parts-crossreference.com

500K+ parts across agriculture, construction, industry, forklifts. Brands: FAG, SKF, Bobcat, ACDelco, Dana Spicer, John Deere, ZF, etc.

✅ Uses headless browser (Puppeteer) to render Livewire content. Extracts part numbers, types, weights, and alternative counts.

Brands (comma-separated):

Max/brand:

Import CSV Data

Bulk import bearing data from a CSV file. Required column: partnumber. Optional: manufacturer, type, inner_diameter, outer_diameter, width, seal_type, replaces.

Source Name *

Source URL (optional)

CSV File *

Seed Sample Data

Load sample bearing data for testing the cross-reference engine.

Process Raw Parts

Run normalization pipeline on pending raw parts.

Link Equivalences

Second pass: create cross-reference edges between normalized parts using interchange data. Run this after processing raw parts.

Current equivalence edges: 260254

Run All Pipeline

One-click pipeline: Normalize all raw parts → Link all equivalences. This may take several minutes for large datasets.

Pending Raw Parts

172925

Normalized Parts

260254

Equivalence Edges

Data Sources

Registered sources for part data ingestion.