Six-figure · Other round · Software · Vienna, Austria
Vienna-based Deepentix, which builds auditable enterprise AI for regulated industries, has raised a six-figure pre-seed SAFE from tecnet equity, Rockstart and business angel Kambis Kohansal Vajargah. The company says the capital will fund research and development and the expansion of initial pilot projects.
Deepentix and its investors described the financing as six-figure but did not disclose the exact amount or currency. The round was completed through a SAFE, an agreement that can convert into equity later.
Turning traceability into a product layer
Deepentix is targeting teams in pharmaceuticals, medical technology, insurance and banking that work across studies, patents, standards and internal reports. Its platform converts those documents into knowledge graphs so an AI-generated answer can be traced to the exact sentence behind it, according to the investor announcement. Rather than delivering a bespoke consulting project, Deepentix wants customers to create, operate and adapt those graphs themselves.
The commercial mechanism is control as much as accuracy. Deepentix says it combines specialised small language models with a structured evidence layer and can run on a customer's own infrastructure, including private-cloud and on-premise environments. That design is meant to keep sensitive data inside the buyer's chosen boundary while giving compliance, risk and scientific teams a reviewable chain of evidence. If the workflow holds up in production, the product can become part of how those teams approve decisions rather than another general-purpose assistant sitting beside their systems.
That positioning also defines the next execution test. Regulated buyers will need more than a convincing demo: integrations must preserve document hierarchies, reviewers must be able to validate extracted logic, and cited answers must remain reliable as source material changes. The planned pilots matter because they can turn Deepentix's technical claims into evidence about deployment time, review effort and repeatable customer value.
Early capital for an enterprise proof point
The company is still at the pilot-building stage, so the financing is not evidence of broad deployment. It gives the team room to refine the platform and test whether a self-serve knowledge-graph product can replace part of the custom work usually needed to make enterprise AI auditable. Tecnet equity said it backed the founders' mix of scientific and international-company experience; Rockstart lists Deepentix in its 2026 Emerging Tech portfolio.
Explore this theme
The Infrastructure Test (Part I) — Alessia Russo, EUVC · 21 April 2026. A practical framework for asking whether vertical AI creates structural advantage or leaves the underlying infrastructure provider in control.
