About CyberNeuron
A collective of ML scientists and engineers asking one question: what if AI could understand the world the way living systems do?
Intelligence That Understands
CyberNeuron is an AI research lab. We take on machine learning problems that don't have a known solution yet, and we work them through to a system that runs in production.
The problems that interest us are the ones where the hard part is measurement. What does emotion in a vocalization actually look like in the data? How do you tell an expert's question from a novice's? Answering that kind of question means building the dataset before the model, and working closely with people who know the domain far better than we do.
WoofSense came out of that process. So did our work on adaptive intelligence and on vision. They are what the lab has built so far, and we expect the list to keep growing.
Our Values
Rigor
We approach every problem with the discipline of peer-level research. Our models are trained and evaluated on real-world data. Our claims are backed by evidence. We don't ship things we don't believe in.
Impact
Research without application is incomplete. Every project we pursue has a clear path from model to product to real-world outcome — whether that's a dog owner understanding their pet better, or an analyst making a more informed decision.
Openness
We collaborate. With academia, with industry, with domain experts whose knowledge we need to build AI that actually works in the real world. The best AI is built at the intersection of machine learning and deep subject matter expertise.
Our Team
Menua Gevorgyan
Menua has spent 10+ years building machine-learning systems across domains — signal processing, behavioral AI, and applied research — with a PhD in quantum optics behind the measurement thinking. He founded CyberNeuron to turn hard research problems into systems that actually ship.
Razmik Hovhannisyan
Razmik leads scientific validation and research rigor at CyberNeuron. A computational and experimental physicist (PhD, Stockholm University), he holds an MSc and BSc, both summa cum laude, from MIPT. His research spans algorithms, inverse problems, and numerical modeling, with first-author work in Physical Review Letters, Nature Communications, and Nano Letters. At CyberNeuron he owns experimental design and statistical rigor — making sure every claim the models make holds up against real data.
Based in Yerevan, Armenia
CyberNeuron is headquartered at Anahit Techcenter in Yerevan — part of Armenia's growing technology ecosystem. Yerevan has established itself as one of the most active tech hubs in the Caucasus and Eastern Europe, with a strong tradition of deep technical education and a growing concentration of engineering talent.
Our location gives us access to exceptional ML and engineering talent, proximity to European research networks, and the focus that comes from working outside the noise of Silicon Valley.
3 Hakob Hakobyan St
Yerevan, Armenia
Who We Work With
CyberNeuron partners with organizations across the technology ecosystem — from early-stage startups to established enterprises and academic institutions.
DeepTech Startups
We work with early-stage and growth-stage technology companies integrating AI into core products. Whether you need a behavioral API, a custom model, or a research partnership — we move fast and build for production.
Enterprise & Corporations
For organizations deploying AI at scale, we offer custom model development, research partnerships, and technical advisory engagements grounded in peer-level ML expertise.
Universities & Research Institutions
We actively collaborate with academic research groups — sharing datasets, co-authoring work, and providing model access for research use. Advances in behavioral AI and computer vision benefit from open scientific exchange.
Let's Work Together
Whether you're a researcher, a product team, or an enterprise — we'd love to explore what's possible together.