SpecMake

TDS to JSON

Convert Technical Data Sheets to Structured JSON

Technical Data Sheets contain the product specifications that PIM systems, ERP databases, and e-commerce platforms need. But the data is locked in PDF tables, multi-column layouts, and formatted text that generic extractors can't parse reliably.

SpecMake reads your TDS — PDF or DOCX, any language — and extracts every property-value-unit combination into clean, structured JSON. Download directly, or export as Excel, JSON-LD, PDF, or DOCX. The extraction handles multi-row tables, nested properties, conditional values, and domain-specific symbols.

Structured JSON output

Every property extracted as a key-value-unit triple. Section headings preserved. Ready for PIM import, database ingestion, or API consumption.

Excel export included

Not just JSON — download as Excel with source and translated values side by side. One spreadsheet per document, ready for review or bulk import.

Domain-aware extraction

Auto-detects your industry (construction, hydraulics, coatings, etc.) and applies the correct terminology mapping. No manual configuration needed.

JSON-LD for DPP compliance

Export as schema.org Product JSON-LD — the machine-readable format that Digital Product Passport registries consume.

Example: Hydraulic valve TDS → JSON

Source PDF table

Max. operating pressure     350 bar

Flow rate                   40–120 L/min

Operating temperature       -20 to +80 °C

Weight                       3.8 kg

Structured JSON output

{
  "heading": "Specifications",
  "fields": [
    {
      "key": "Max. operating pressure",
      "value": "350",
      "unit": "bar"
    },
    {
      "key": "Flow rate",
      "value": "40–120",
      "unit": "L/min"
    }
  ]
}

Why generic PDF extractors fail on technical data sheets

Generic PDF-to-JSON tools treat every document the same. They extract raw text or table grids without understanding what the data represents. For a technical data sheet, this means:

Table columns get misaligned — a value in one row ends up paired with a property name from another row
Units are separated from values — "350" and "bar" become independent text fragments instead of a linked pair
Conditional values and ranges are flattened — "40–120 L/min at 20°C" loses the condition context
Symbols and special characters (●, ○, ±, ≥) are dropped or corrupted
Multi-section documents get merged into a flat list, losing the section structure that gives fields meaning

SpecMake's extraction pipeline uses position-aware text analysis alongside AI understanding. It reconstructs table rows from Y-coordinates, identifies column boundaries from X-gaps, and detects vector-drawn symbols from the PDF operator list. The AI then structures the extracted text into property-value-unit triples, preserving section hierarchy and conditional relationships.

From JSON to PIM, ERP, and e-commerce

Once your TDS data is in structured JSON, it can flow into any system that consumes product data:

PIM systems (Akeneo, Salsify, inRiver, Pimcore) — map JSON fields directly to product attributes
ERP systems — import product specifications into material master records
E-commerce platforms — populate product pages with structured technical data
Product configurators — use extracted parameters for compatibility checks
DPP registries — export as JSON-LD for Digital Product Passport compliance

For a deeper look at how extracted data maps to PIM schemas, see our spec sheet to PIM integration guide.

Works across all industrial verticals

The extraction pipeline auto-detects the document's industry domain and applies the correct terminology mapping. Whether your TDS describes a hydraulic valve, an insulation panel, a coating system, or a food-grade pump — the same upload workflow produces structured output with domain-appropriate field names and units.

13 industrial domains supported, from construction materials and coatings to hydraulics and food processing.

Related articles

How to Extract Data from Spec Sheets

Why generic extractors fail on technical documents, and how AI-powered extraction works.

From Spec Sheet to PIM

Automating the gap between paper spec sheets and PIM systems. JSON as the bridge.

Spec Sheet to Excel

Turn any technical PDF into a clean spreadsheet. No copy-paste, no manual cleanup.

DPP-Ready Export

JSON-LD export mapped to schema.org — the format Digital Product Passport registries consume.

Convert your first TDS to JSON

Upload a technical data sheet (PDF or DOCX) and download structured JSON, Excel, or JSON-LD in seconds.

Upload a document — it's free

No credit card required. Your first document is free.