75F Building Automation
AI-enabled building automation combining connected sensors, controllers, and cloud operation.
Pricing to verifyCustom quote; amount unconfirmed
Productivity & Research
eScriptorium
Open-source historical text recognition with trainable Kraken models.
What it does
eScriptorium manages image import, layout segmentation, automatic transcription, manual correction, and model training around the Kraken engine. Institutions can run their own instances and export text, layout information, and trained models.
Who it is for
Research teams needing control over historical recognition models and infrastructure.
What stands out
- Kraken-based segmentation and handwriting recognition
- Line/image editing and custom model fine-tuning
- ALTO/PAGE XML export and downloadable models
Limitations to consider
- No unrestricted public instance is hosted by the core team.
- Hosting, compute, and administration require resources.
Integrations & connections
Tasks it can help with
Sources & verification
This profile is based on vendor documentation, checked Oct 22, 2026. It is not a hands-on review or a performance benchmark. Unknown details are marked explicitly.
Current historical-text workflow overviewOpen-source deployment and institutional access policiesCustom training, editing, and complex scriptsStructured export and model portabilityOfficial Kraken-based workflow documentationSuggest a correction