The BrainAccess team presented our paper on Jena, an AI copilot for EEG acquisition that pairs deterministic safety rules with LLM reasoning, at the Industry Track of RuleML+RR 2026 in Vilnius. This post covers the conference and what Jena does.
From 24-26 August 2026, the BrainAccess team joined RuleML+RR 2026, the 10th International Joint Conference on Rules and Reasoning, right here in our home city of Vilnius. The conference was the centerpiece of Declarative AI 2026 and brought together researchers and practitioners working on rule-based reasoning, knowledge representation, and declarative approaches to AI.
RuleML+RR has spent a decade building bridges between academic research and industrial applications in semantic reasoning. Its Industry Track exists specifically for that purpose: real-world experiences of applying rules and reasoning technologies in practice, spanning topics like neuro-symbolic and generative AI integration, LLM governance and verification with rules, and agentic systems.
Our contribution: Jena AI
We were proud to present “Jena: Combining Deterministic Rules and LLM Reasoning for Grounded EEG Acquisition Support”, authored by Martina Berto, Linas Raičinskis, and Aleksandras Voicikas, and accepted for the Industry Track.
Jena is an AI-assisted component integrated into BrainAccess Board, our EEG acquisition software. Rather than acting as a general-purpose chatbot, Jena works as a constrained acquisition copilot: it helps with device setup, troubleshooting, live signal quality monitoring, and results interpretation, while staying grounded in curated manufacturer knowledge and live device state.
💡 Jena in a nutshell
The core idea is architectural. A local plugin captures real-time context (i.e., connected devices, recording state, impedance readings, EEG windows) and sends it to a server backend. The backend combines this live context with an LLM-guided navigation system over curated, allowlisted documentation, plus a set of deterministic rules that govern what the model can access, claim, and expose. The result is an assistant that reasons flexibly in natural language, but can’t hallucinate signal quality, invent device actions, or step outside its scope. Every claim is tied to actual measurements and explicit system state.
Jena currently support setup guidance, hardware and software troubleshooting, signal quality feedbacks, and hyperscanning!
For a conference dedicated to combining rules and reasoning, Jena offered a concrete, deployed example of that philosophy applied to physiological hardware, a domain where trust, auditability, and safety guardrails really matter.
Try Jena AI for Free
Discover our EEG-aware assistant, integrated in BrainAccess Board. Free in beta.
The full experience
Beyond the talks and papers, the event was every bit as rich socially as it was intellectually. We met researchers and practitioners working on everything, from cutting-edge applications to advanced theoretical reasoning, and traded ideas late into the evening. The organizers also treated us to a proper Vilnius send-off: a social dinner up the Vilnius TV Tower, where a few of us worked up the nerve for the “Edge Walk”: harnessed and guided, we walked along an open platform some 160 meters up, quite literally on the edge of the tower! We capped it off with sunset and dinner at the tower’s revolving restaurant, with a special view of our beloved Vilnius from above.
It was a great few days of talks, discussion, and networking with the rules-and-reasoning community, and an opportunity to meet new people and catch up with previous team members!
Next year the event will be in Vienna! Follow the organizers for more details.
