The Irish Building Stock Observatory links building performance with social, economic and financial data so retrofit investment can be targeted with evidence.

An ethical AI research learning platform that makes the student research process visible, supporting academic integrity rather than policing outcomes.

Takes ProACT, a validated digital platform for managing multimorbidity, from research prototype to a commercial product for community-based care.

Verified, context-aware name pronunciation infrastructure, combining owner-recorded audio with linguistic validation for enterprise-scale use.

A RegTech platform for reporting serious AI incidents under Article 73 of the EU AI Act, connecting enterprise compliance systems directly to regulators.

An AI platform recommending personalised prosthesis configurations from patient data, improving socket fit, comfort and rehabilitation outcomes.

Predicts the energy a computing workload will use before it runs, helping data centres cut costs, reduce emissions and meet EU reporting rules.

A copilot for product definition, helping product managers turn scattered, ambiguous inputs into clear, consistent requirements before development starts.

Translates complex grid planning simulations into visuals non-technical decision-makers can act on, supporting multi-million euro infrastructure decisions.

A game-based AI language learning platform combining speech, text and image generation to improve vocabulary, literacy and learner motivation.

Detects combustion drift, fouling and steam trap failure in SME thermal plants, recovering fuel losses of €14,000–€25,000 a year without costly sensors.

A no-code fault detection platform that lets maintenance engineers deploy predictive AI without data science expertise or large training datasets.

Creates privacy-preserving synthetic knowledge graphs that retain the relationships and semantic context of real data, unlocking restricted datasets for AI.

Predicts optimal AV1 encoder settings in a single pass, removing the repeated trial encodes that drive up cloud costs and energy use.

A multi-agent AI platform that applies each company’s own terminology and style rules to machine translation, cutting costly human post-editing.

A multilingual conversational platform explaining conditions and treatments using only regulator-approved content, adapted to each patient’s literacy level.

An AI-enabled platform that prepares people with serious illness for shared decision-making, turning advance care planning from paperwork into conversation.

Certifies metal powder oxidation at the point of use in additive manufacturing, replacing destructive lab tests that take three to five days.

End-to-end encrypted document collaboration where keys never leave the browser, so no provider can read or be compelled to disclose your content.

An AI pedagogical engine for language education that embeds task-based teaching and CEFR-aligned progression, not just AI-generated content.

An accurate, low-cost dust monitoring system built for small construction firms, making occupational exposure compliance affordable on site.

High-bandwidth optical neuromorphic AI chips that cut the electricity and water data centres consume to run rapidly growing AI workloads.

Verifies authentic student authorship by capturing the writing process itself, rather than analysing finished text for AI-generated content.

Keeps CCTV footage analysable while protecting identities, offering reversible privacy rather than permanent anonymisation or blanket encryption.

Turns unstructured patient complaints and incident reports into structured, evidence-linked insights, so emerging safety risks surface earlier.

An AI genomic prediction engine that learns directly from complex, noisy datasets to deliver reliable trait predictions across crops and environments.

An EU policy intelligence platform with a natural language interface, giving public affairs teams strategic insight into positions and stakeholder networks.

Parametric climate risk products for agriculture and renewable energy, helping farmers, lenders and operators manage weather-driven revenue volatility.

An AI-assisted digital twin that automates the design, configuration and validation of converged 5G and Time-Sensitive Networking for industrial automation.

Relieves congestion on hospital wireless networks by enabling direct device-to-device communication, removing access points as a bottleneck.

Uses holographic volumetric 3D printing to produce next-generation intraocular lenses, improving visual quality after cataract surgery.

Runs AI inference locally by optimising GPU memory, letting organisations deploy multi-agent systems without cloud API costs or data leaving the building.

The first hearing device to improve speech comprehension by shifting where sound appears to come from, developed for autistic listeners in noisy settings.

A chemical-free precision weeding, fertigation and irrigation platform for organic farms, using computer vision to target hot water at individual weeds.

Processes the 60 million-plus thermal data points a metal 3D print generates, predicting density and hardness without costly ex-situ testing.

A no-code platform for building AI, machine learning and analytics pipelines by drag-and-drop, from school classrooms to research teams.

An AI platform for investigating cryptocurrency crime, linking transactions across wallets and consolidating victim reports for law enforcement.

A low-cost, disposable RFID tag that signals remotely if frozen goods rise above zero, enabling continuous cold-chain monitoring across the supply chain.

Protects UAV communications from jamming by fusing spectrum measurements, RF fingerprinting and terrain data to reroute flights in real time.

A game-based platform helping primary pupils practise maths while reducing maths anxiety, with adaptive gameplay and progress tools for teachers.

A machine translation system rendering English into Irish Sign Language through an avatar, extending language technology to a long-excluded community.

A channel estimation technique that cuts signalling overhead in massive MIMO to a single pilot sample, improving spectral efficiency for 6G networks.

A mobile app that interprets sign language into English in real time, adapting to the signer’s fluency and dialect across everyday interactions.

Low-cost Doppler radar sensors that identify and count flying pollinators non-invasively, bringing biodiversity monitoring into environmental IoT networks.

Millimetre-wave radar that monitors breathing and micro-movements through clothing and barriers, with no wearable or physical contact required.

A platform merging motion capture with motion synthesis, letting technical and non-technical users create expressive, believable animated characters.

An AI co-pilot for educators that turns multimodal learning data into conversational feedback, revealing learning gaps across whole cohorts.

An EV fleet battery platform using real-time vehicle data and AI to extend battery life, cut maintenance costs and improve fleet reliability.

Connects GPS and accelerometer data with tactical video analysis, automating insight that until now required labour-intensive manual coding.

A remote cardiac rehabilitation platform combining wearables and validated exercise protocols, reaching patients unlikely to attend centre-based programmes.

A privacy-respecting wearable platform turning everyday activity data into early warnings, supporting independent living with chronic conditions.

A federated learning framework that detects ransomware across healthcare networks without sharing patient data or relying on known signatures.

Reinforcement learning xApps that let Open RAN networks adapt to new tasks without forgetting previous ones, supporting diverse 5G service demands.

Optimises how 6G networks allocate computing resources at the edge, so demanding services like autonomous vehicles run reliably on constrained hardware.

Generates photorealistic 3D avatars from ordinary 2D images, enabling holographic telepresence without specialist capture hardware.

A governance platform that maps every AI model a company runs, tracks performance and cost, and recommends where models can be consolidated.

Automatically assesses road surface condition from imagery using computer vision, replacing slow manual inspection of defects such as potholes.