Independent AI Research Lab
Independent fundamental and applied research at the intersection of cognitive architectures, AI safety in public spaces, and applied AI for knowledge work. Self-funded, open access, with civil-liberties architecture as a design principle.
Architecture
ZenBrain as the cognitive core, modular applications for different domains. The same seven-layer memory architecture carries every application — only the domain adaptation varies.
Memory layers
Three research tracks
How AI remembers, reasons, and learns over time. Seven-layer memory architecture (ZenBrain), sleep consolidation, neuromodulator-driven consolidation, FSRS-driven review.
Research outputs
Research on the early detection of critical situations in public spaces — explicitly without biometric identification. Three-outlier model, helper protection as an objective, civil-liberties architecture in eight mandatory corrections.
Research outputs
AI-augmented meetings, project management, mediation, and process mapping. How organisations structure AI adoption — step by step, with humans at the centre.
Research outputs
Research principles
Research outputs are published on arXiv and Zenodo with DOI — freely accessible, permanently archived.
Core research documents are anchored in time via OpenTimestamps. This produces a forensically verifiable priority anchor — independent of later publication dates.
Safety research with fundamental-rights grounding — explicitly without biometrics, with helper protection and the safeguard of political assemblies as structural architecture decisions.
Algorithms are not only published — they run in production systems. Research has to carry.
Methodological standards
Operational standards against which our research holds itself accountable — from the first hypothesis to publication.
Research questions, hypotheses, and architectural decisions are documented before the work begins and anchored in time via OpenTimestamps. This produces a forensically verifiable priority record — independent of the later publication date.
Algorithm source code and test suites are public (github.com/zensation-ai). Datasets, eval scripts, and configurations are released with each publication so that third parties can replicate our results.
Safety-relevant components are reviewed externally before publication — via Notified-Body pre-assessments (TÜV SÜD, TÜV Rheinland, Bureau Veritas) and legal opinions on the fundamental-rights conformity of the architecture in question.
Failed hypotheses, negative findings, and ablation studies are published as well. Methodological transparency that includes unfavourable results is what makes replication meaningful in the first place.
Research data is collected in the sense of Art. 89 GDPR in conjunction with § 27 BDSG such that no biometric features are produced. Annotation schemes are explicitly designed for the minimum that supports the respective research question.
Origins
A three-year arc — from the first conceptual sketch to architectural consolidation to open-access publication. The waypoints below are documented in the internal development repository and in the publications themselves.
Run-up. Algorithmic work on statistical models, deterministic decision systems, and real-time processing. In parallel, several years of digitalisation and transformation projects — a twofold preparation that later shows in the architecture.
The idea for a cognitive architecture with real long-term memory takes its first form. First sketches of a multi-layer memory structure drawing on neuroscience (Ebbinghaus, FSRS, Hebbian learning).
Architectural principles take concrete shape: seven memory layers, sleep consolidation as a methodological core, Bayesian confidence propagation. In parallel, the civil-liberties architecture for the public-safety track matures.
First productive implementations of the memory layers. Build-out of the test suite, methodological pre-registration via OpenTimestamps. Beginning of the research documentation as an independent source.
Start of the productive implementation phase in the current internal development repository. Eight consecutive architecture phases in the first two weeks.
Consolidation as an open-source project. ZenBrain is extracted as a standalone package family and released on npm under Apache 2.0.
Academic preparation. LaTeX paper infrastructure, ORCID profile, arXiv endorsement process, academic reputation strategy.
ZenBrain preprint on arXiv (2604.23878) and Zenodo (DOI 10.5281/zenodo.19353663) — open access, fully documented, with replication material.
Preparation of follow-up publications for peer-reviewed venues. Exploratory exchanges with universities and research consortia across the European research area.
Research agenda
Exploratory work and preparatory steps — kept in the subjunctive throughout until written consent or binding funding decisions are in place.
Horizon: 2026 – 2027
Concrete dates may shift; names of cooperating individuals and institutions are disclosed only after written consent.
Publications and identifiers
Open-access publications, software releases, and academic profiles at a glance.
Collaboration
We regularly explore collaborations with universities, research institutes, funding agencies, and public authorities. For research or consortium inquiries, please get in touch.
Send research inquiry →More from this research
Three tracks, one core, a shared ethics foundation.