VERTIKAL AI APS

Their core work

Vertikal AI is a Danish technology SME that builds applied AI systems for industrial environments, with a clear focus on machine learning for predictive operations and IoT data pipelines. Their PAVIMON project applied AI to predict failures in wind turbines, while AI-TRAIN tackled the harder problem of transfer learning — enabling AI models trained in one industrial context to work in another without full retraining. Both projects were coordinated solo, which is consistent with a small product or consulting firm packaging a specific technical capability into a validated business case for EU funding. They sit at the intersection of industrial AI and operational efficiency, targeting sectors where unplanned downtime is expensive.

What they specialise in

How they've shifted over time

Vertikal AI's two projects span only 2019–2021, so there is no long-term trajectory to analyze. Within that narrow window, there is a discernible step up in ambition: PAVIMON addressed a specific vertical (wind turbine maintenance), while AI-TRAIN generalized the problem — using transfer learning to make AI models portable across industrial settings. This suggests a deliberate move from domain-specific application toward foundational AI methodology that can serve multiple sectors. Given the small timeframe and data, this reads more as a product roadmap progression than a strategic pivot.

They appear to be moving toward generalizable AI infrastructure for industry — the kind of capability that could underpin multiple verticals — rather than staying locked to a single application domain like wind turbines.

How they like to work

Both H2020 projects were coordinated by Vertikal AI with no recorded consortium partners, which is typical for SME Instrument Phase 1 grants where a single company validates a business concept solo. This means there is no evidence of how they behave in multi-partner settings — they have not yet been tested as consortium members or in collaborative R&D. For anyone considering them as a partner, they bring technical AI capability but come without the consortium track record that larger or more experienced SMEs have.

No consortium partners are recorded across either project, consistent with solo SME Instrument applications. Their collaborative network in the H2020 system is effectively zero, with no cross-border partnerships on record.

What sets them apart

Vertikal AI occupies a narrow but commercially interesting niche: applied AI for industrial operations, with demonstrated EU validation through two funded projects in under two years. Most SMEs in Denmark working on industrial AI are either hardware-adjacent (sensors, PLCs) or large-firm spinouts — a lean AI-native company with H2020 credentials in both energy and IoT is relatively uncommon at this size. Their transfer learning focus is technically specific enough to be credible, and wind turbine predictive maintenance is a well-defined market with real buyers.

Highlights from their portfolio

• AI-TRAIN
  The largest of their two projects (EUR 117,000) and the broader in scope, AI-TRAIN's focus on transfer learning for Industrial IoT touches a commercially significant challenge — making AI models reusable across industrial settings without expensive retraining.
• PAVIMON
  An early-stage feasibility project (SME Instrument Phase 1, EUR 50,000) validating AI-driven predictive maintenance on wind turbines — a high-value application in an energy sector actively seeking to reduce O&M costs.