The science behind c:node.
The c:node Graph is the intelligence in c:node: a knowledge-graph AI that binds every statement to a source (technical term: neuro-symbolic). It builds on open models (open source) and grows steadily with every case. It is built to work traceably instead of guessing — every statement bound to a source, every step auditable. This article summarises the peer-reviewed evidence behind it: why pure language models aren't enough when it matters, why knowledge graphs + rules close the gap, and how this fits the requirements of the EU AI Act.
1 Why pure language models aren't enough when it matters.
Large language models form sentences by predicting the statistically most likely next word. The US standards institute NIST states that fabricated statements (confabulations) are inherent to the design of such models — not a slip, but a property of the architecture.1 Looking things up in real documents (retrieval-augmented generation) lowers the error rate significantly, but does not eliminate it: even with the right documents, simple models still add unevidenced content.5 For decisions with consequences, “mostly right” isn't enough.
2 The way out: recognise patterns and check them against rules.
Research calls the approach neuro-symbolic AI: the pattern recognition of neural networks is combined with the auditability of symbolic rules and knowledge graphs.2 A peer-reviewed survey (NAACL 2024) shows that knowledge graphs used as an external knowledge source reduce hallucinations and increase reasoning accuracy.3 And in multi-step reasoning — combining several linked facts into one answer — the graph-based variant is demonstrably superior to plain text search.4 That is exactly what c:node is built on.
3 Traceability isn't optional — it's mandatory.
For high-risk AI, the EU AI Act requires exactly what this architecture delivers anyway: complete logging (Art. 12), transparency and traceable documentation (Art. 13) and the possibility of human oversight (Art. 14).6 Because c:node backs every statement with source and history (provenance), every decision remains auditable afterwards — without the inner workings of the model having to be disclosed.
4 How the c:node Graph is built — and why it grows.
The c:node Graph builds on open models (open source) — interchangeable without changing the traceable logic —, runs in the cloud, on your servers or fully offline and so stays sovereign within your environment. It isn't a static model, but a living knowledge graph that grows with every case, per tenant. The division of labour: you steer with c:node, the c:node Graph provides the evidence, the agents do the work.
5 Scientific advisory board.
The direction of c:node is supported by academic expertise in decision research and AI.
Assoc. Prof. Dr. Tuna Çakar
Research on artificial intelligence, decision processes, machine learning and applied neuroscience — the cognitive and methodological foundations of robust decision intelligence. Selected works:
- Physicians' ethical concerns about artificial intelligence in medicine — Frontiers in Public Health, 2024.
- Unraveling neural pathways of political engagement: bridging neuromarketing and political science — Frontiers in Human Neuroscience, 2023.
- An investigation of the neural correlates of purchase behavior through fNIRS — European Journal of Marketing, 2018.
Sources
- NIST, Artificial Intelligence Risk Management Framework: Generative AI Profile (NIST AI 600-1), 2024 — confabulations are inherent to the design of generative models. nvlpubs.nist.gov…
- Hitzler et al. (eds.), Handbook on Neuro-symbolic AI and Knowledge Graphs, IOS Press, 2025 — knowledge graphs as a bridge between rules and learning models. iospress.nl…
- Agrawal et al., “Can Knowledge Graphs Reduce Hallucinations in LLMs? A Survey”, NAACL 2024 — KGs as an external knowledge source reduce hallucinations and increase reasoning accuracy. aclanthology.org…
- GraphRAG-Bench — graph-based AI outperforms plain text search in multi-step reasoning. arXiv:2506.02404, 2025. arxiv.org…
- MEGA-RAG (Multi-Evidence RAG) — retrieval reduces fabricated statements by more than 40%, but does not eliminate them. PMC12540348, 2025. ncbi.nlm.nih.gov…
- European Union — EU AI Act, High-level Summary (Art. 12 logging, Art. 13 transparency, Art. 14 human oversight). artificialintelligenceact.eu…
- Assoc. Prof. Dr. Tuna Çakar — publication profile, Google Scholar (as of 2026). scholar.google.com…
The sources support the underlying principles (neuro-symbolic AI, knowledge graphs against hallucination, EU AI Act requirements). Details on the scientific advisory board are public (Google Scholar); R&D key data will be added after internal approval.