SPHINX developed a cyber security toolkit for healthcare using AI, honeypots, and homomorphic encryption; LAW-GAME applied AI to security training.
AIDEAS OU
Estonian AI company applying machine learning, cybersecurity tools, and VR training systems across security, healthcare, and manufacturing sectors.
Their core work
AIDEAS is an Estonian technology SME specializing in artificial intelligence applications across security, healthcare, and manufacturing domains. They build AI-driven tools including cybersecurity toolkits, explainable AI for manufacturing decision support, and immersive training platforms using serious games and virtual reality. Their work spans from protecting health-sector IT infrastructure with advanced encryption and honeypots to developing graph-based machine learning systems for industrial use.
What they specialise in
XMANAI focused specifically on explainable AI, graph deep learning, and hybrid ML for manufacturing decision support.
LAW-GAME built interactive digital gamification for experiential training; Search and Rescue developed technologies for first responder crisis scenarios.
Carbo4Power (offshore turbine blades) and iclimabuilt (climate-adaptive building materials) suggest AIDEAS contributes digital or AI components to physical-domain projects.
Search and Rescue targeted early victim location under collapsed structures; LAW-GAME addressed law enforcement training through gamification.
How they've shifted over time
AIDEAS entered H2020 in 2019 with a strong cybersecurity focus — blockchain, homomorphic encryption, honeypots, and vulnerability assessment for the healthcare sector (SPHINX). By 2020-2021, their focus shifted decisively toward explainable AI, graph-based machine learning, and immersive training via serious games and VR. This evolution shows a company moving from defensive security tooling toward broader AI explainability and human-centric training applications.
AIDEAS is moving toward human-interpretable AI and experiential training platforms, suggesting future projects will combine explainable ML with simulation-based learning environments.
How they like to work
AIDEAS operates exclusively as a consortium participant — they have never coordinated a project, which is typical for a specialized SME contributing focused technical modules. With 111 unique partners across 21 countries in just 6 projects, they join large research consortia (averaging ~19 partners per project) and appear comfortable working with diverse, multinational teams. This breadth of partnerships without repeat patterns suggests they are valued for specific AI/software capabilities rather than long-standing bilateral relationships.
Despite being a small Estonian company, AIDEAS has built connections with 111 distinct partners across 21 countries through large EU consortia. Their network is geographically broad across Europe with no obvious concentration in a single region.
What sets them apart
AIDEAS bridges the gap between advanced AI research and practical application domains like security, healthcare, and manufacturing — a combination rarely found in a single SME. Their ability to contribute AI components (from graph deep learning to VR training) to projects in vastly different sectors makes them a versatile technology partner. For consortium builders, they offer an SME-sized, agile AI team based in Estonia with proven experience delivering across multiple EU project types.
Highlights from their portfolio
- LAW-GAMETheir largest funded project (EUR 484,625), combining AI, serious games, and VR for law enforcement training — an unusual intersection of gaming and security.
- XMANAIFocused on explainable AI for manufacturing, featuring graph deep learning and hybrid ML — reflects their push into interpretable AI systems.
- SPHINXTheir first H2020 project, building a cybersecurity toolkit specifically for healthcare using homomorphic encryption and AI — an early signal of their cross-domain AI approach.