Europe Day and the European Year of Skills: why the EIC is spotlighting applied AI

On Europe Day, 9 May, and in the context of the European Year of Skills, the European Innovation Council community highlighted a set of EIC-backed companies that are applying artificial intelligence to improve business processes and digital capabilities across the EU. The companies discussed offer different technical approaches to the same policy objective. They position their tools as ways to raise productivity, reduce low-value tasks and expand access to digital tools for non-expert users.

The European Innovation Council supports startups and scaleups with funding and business acceleration. The EIC plays a role in advancing technologies that align with EU priorities such as digital transformation and skills development. The three companies featured here illustrate how EIC-funded projects translate research and R&D into commercial products. Their claims are notable but merit scrutiny on practicality, governance and measurable impact at scale.

Three approaches: edge analytics, visual support and no-code NLU for automation

Stream Analyze: pushing AI onto the device for real-time analytics

Stream Analyze, a Swedish company, is developing an end-to-end edge AI platform called sa.engine. The platform aims to let data scientists, engineers and domain experts with limited coding experience build and run advanced real-time analytics and AI models on microcontrollers and other edge devices. The company frames its approach as an alternative to cloud-dependent architectures that move large volumes of data to central servers for processing.

Stream Analyze says this approach enables real-time analysis without uninterrupted connectivity and reduces the storage and compute footprint compared with cloud processing. The company reports use cases including quality inspection in automotive production, predictive maintenance for mining loaders and vehicle fleet monitoring in collaboration with partners such as Volvo Trucks and Boliden.

Jan Nilsson, Stream Analyze CEO, emphasised a customer-first posture and the company’s belief that Europe must invest in hands-on training and inclusive initiatives to raise digital skills. He said, "Our primary focus is to empower our customers to harness the power of AI by making their devices, machines, and assets smarter." He recommended education and training that prioritise practical learning and inclusivity to build a resilient workforce.

BlinkIn: visual support, computer vision and augmented reality for user guidance

BlinkIn is a German startup building multimodal customer support and visual assistant technology that works on standard smartphones without extra hardware. EIC funding supports further development. BlinkIn aims to combine computer vision, AI-driven diagnostics and augmented reality to guide users interactively through setup, troubleshooting and maintenance tasks.

BlinkIn plans to develop bots that automatically analyse visual inputs, generate AI or machine learning recommendations and then lead users step by step using AR. Founder Josef Süß said the company wants to provide immersive experiences while respecting user privacy and ethical considerations. He recommended three policy or ecosystem actions to lift digital skills: invest in digital literacy programs, foster cross-industry collaboration and encourage lifelong learning.

DeepOpinion: no-code NLU and agentic automation for document-heavy processes

DeepOpinion, which the EIC has funded, positions itself as a B2B SaaS provider of no-code tools to automate workflows that are dominated by unstructured text and documents. Its AutoNLU engine and related components aim to automatically train high-performance natural language understanding models across many languages and to integrate with robotic process automation and existing enterprise systems.

DeepOpinion says its proprietary ML engine was developed over four years and supports automated model selection and training for 100 plus languages. CEO Stefan Engl argued for rethinking STEM education to focus on real-world problem solving and lifelong learning. He described a vision where knowledge workers gain "an AI Co-worker" to offload boring tasks.

The company also highlights recent trends such as generative reasoning and zero-shot capabilities where large models handle new tasks without task-specific training. DeepOpinion claims adoption by major clients and says its tools improve straight-through processing rates and reduce manual effort in document-intensive industries.

Reality check: technical, regulatory and skills challenges

The three companies illustrate complementary technical strategies but also reveal common hurdles for translating prototype impact into widespread improvements in EU digital skills. First, deployment remains non-trivial. Edge AI relieves bandwidth but requires hardware compatibility, model optimisation and field maintenance. Visual support works where camera access and lighting permit it but depends on reliable CV models and clear privacy rules about video data. No-code NLU lowers the barrier for adoption but can hide model limitations and require robust human-in-the-loop processes for high-risk decisions.

Second, regulation and trust matter. European data protection rules and emerging AI regulation in the EU aim to protect rights and prescribe obligations for high-risk AI systems. Startups and customers must align solutions with GDPR and with evolving EU AI policy. That means clear data governance, auditable model behaviour and privacy-preserving architectures when personal or sensitive data are involved.

All three founders highlighted education and inclusion as necessary complements to technology. Their recommendations line up with broader EU policy thinking on lifelong learning and the need to ease transitions for workers into more digital roles. Yet public impact depends on measurable programmes that reach smaller firms, regional ecosystems and underrepresented groups.

What to watch next

EIC-backed startups are advancing practical AI that targets business productivity and, in principle, can reduce routine work and expand access to digital tools for non-experts. Policymakers should keep testing funding models that combine R&D support with training and procurement incentives for public sector adoption. Industry should prioritise interoperable standards and responsible data practices to accelerate safe uptake.

For observers and potential adopters, the sensible approach is to pilot with clear metrics for user skill uplift, cost savings and risk management. Success will come from technical robustness, data governance and sustained investment in skills rather than from technology announcements alone.

Acknowledgements and caveats

This article reconstructs an EIC Community feature published for Europe Day on 9 May 2023 and expands on technical and policy context. The companies discussed supplied statements about their work and goals. The information is provided for knowledge sharing and should not be read as official EU policy or as an endorsement of any product claims. Continued independent evaluation is recommended when assessing vendor claims about performance and impact.