Limina is the current name of the company formerly branded Private AI. Its product focuses on detecting and de-identifying sensitive information across text and other media for analytics, sharing, retrieval, or AI workflows. Deployment options include customer-controlled environments in some plans.
De-identification can remove obvious PII without eliminating every contextual or re-identification risk. Accuracy depends on language, domain, entity type, and the cost of both missed and over-redacted content.
In Private AI competitors, Limina is a data-transformation component, which places it upstream of the model, retrieval system, and observability stack that each keep their own confidentiality story.
Built on 2 sources (2 external).
Working out connections…
Sources
Working out the neighbourhood…
Model contributions
Measured by git-blame lines per AI model (41 total).
{"width": 320, "height": 320, "data": {"values": [{"model": "GPT-5", "label": "GPT-5 (90%)", "lines": 37, "share": 0.9024390243902439}, {"model": "Claude Opus 4.8", "label": "Claude Opus 4.8 (7%)", "lines": 3, "share": 0.07317073170731707}, {"model": "Claude Opus 5", "label": "Claude Opus 5 (2%)", "lines": 1, "share": 0.024390243902439025}]}, "mark": {"type": "arc"}, "encoding": {"theta": {"field": "lines", "type": "quantitative"}, "color": {"field": "label", "type": "nominal", "legend": {"title": null, "orient": "right"}}, "tooltip": [{"field": "model", "type": "nominal"}, {"field": "lines", "type": "quantitative"}, {"field": "share", "type": "quantitative", "format": ".1%"}], "order": {"field": "lines", "type": "quantitative", "sort": "descending"}}}