TextHarvester
AI-assisted digitisation of historical records and cultural heritage documents
TextHarvester is a digitisation service designed for archivists, historians, and heritage organisations who need accurate, structured transcriptions from handwritten or typed historical documents. Whether you hold burial registers, estate papers, or other archival collections, TextHarvester transforms physical records into searchable, structured data.
Using a dual AI verification approach — where two independent language models cross-check each other's output — TextHarvester delivers high-confidence transcriptions at scale, reducing manual effort while maintaining the accuracy that heritage work demands.
Document Types
- Burial Registers
Parish and churchyard burial records, including names, dates, ages, and plot references.
- Grave Record Sheets
Field survey forms and grave description sheets from cemetery recording projects.
- Estate Papers
Tenancy records, rent rolls, and estate correspondence from landed estate archives.
- Custom Formats
Any structured historical document type — TextHarvester can be configured for your specific record format.
How It Works
Scanned or photographed documents are processed through a pipeline that extracts, verifies, and structures the data. Outputs are delivered as CSV or JSON, ready for import into databases, heritage portals, or GIS systems.
Dual AI Verification
Two independent AI models transcribe each record and results are cross-checked, flagging discrepancies for human review.
Structured Output
Data is delivered in clean CSV or JSON format, structured to match your existing schema or heritage portal requirements.
Batch Processing
Designed for volume — hundreds or thousands of pages can be processed in a single project run.
Recent Project
St Luke's Church Douglas, Cork
- ~500 burial register pages transcribed and structured
- Dual AI approach: GPT vs Claude cross-verification
- Output delivered as CSV, ready for heritage portal import
- Discrepancies flagged and resolved before final delivery
Get in Touch
If you have a digitisation project in mind — or want to discuss whether TextHarvester is a good fit for your collection — reach out directly.