MACHINE INTELLIGENCE SWEDEN AB

Their core work

Machine Intelligence Sweden AB is a Gothenburg-based technology SME that builds AI and machine learning systems, with a particular focus on connecting knowledge — researchers, technologies, and businesses — through intelligent software. They coordinated ScienceRouter, an AI-powered matchmaking platform designed to route research knowledge to where it creates commercial value, which aligns directly with their company identity. In parallel, they contributed technical expertise to LEGaTO, a large research consortium building low-energy tooling for heterogeneous computing environments (FPGAs, micro-servers, dataflow architectures) — indicating they also have hands-on capability in performance-sensitive ML infrastructure. Their profile suggests a small team that sits at the boundary between applied AI product development and computationally intensive research systems.

What they specialise in

How they've shifted over time

Both H2020 projects ran between 2017 and 2020, giving too narrow a window to observe meaningful multi-year evolution. What can be inferred is a simultaneous dual track: LEGaTO placed them in deep technical computing territory (FPGAs, dataflow programming, heterogeneous hardware), while ScienceRouter — which they coordinated — reflects their core product direction: AI-driven matchmaking between research and commercial actors. The absence of keywords on ScienceRouter in the CORDIS data limits this analysis, but the project title and funding scheme (SME Phase 1, a feasibility grant) strongly suggest it was their own product concept rather than a collaborative research role. The trajectory appears to move from being a technical participant in others' research to developing proprietary AI platforms as a product company.

They appear to be building toward a commercial AI product in the research-to-business matchmaking space, using technical computing work as a foundation — potential collaborators should expect a product-oriented SME rather than a pure research partner.

How they like to work

With only two projects, their collaboration profile is limited but meaningful: they joined a large multi-country RIA consortium (LEGaTO) as a participant, likely contributing ML or software tooling expertise, while also securing and leading their own SME Phase 1 grant (ScienceRouter). This dual mode — contributing specialist skills to bigger consortia while independently driving their own product ideas — is typical of technically strong SMEs who use EU projects both to learn and to de-risk their own product development. Their consortium size is small (9 partners across 6 countries), suggesting they engage selectively rather than broadly.

Their H2020 network spans 9 unique partners across 6 countries, modest in scale but geographically spread across Europe. No repeated partner relationships are visible in the two-project dataset, so no loyalty pattern can be established.

What sets them apart

Machine Intelligence Sweden's differentiating asset is that their own product — an AI matchmaking system for research and innovation — is itself an H2020-validated concept, giving them credibility in exactly the space they serve. Unlike most technical SMEs that only build tools for others, they have experience designing and pitching a knowledge brokering platform from scratch, which is rare. For consortium builders, they bring a specific combination of ML engineering depth (informed by heterogeneous computing research) and applied AI product thinking — useful in projects that need both technical rigor and commercial translation.

Highlights from their portfolio

• ScienceRouter
  The only project they coordinated, and the one closest to their commercial identity — an AI-powered matchmaking system for research knowledge, funded under SME Phase 1, essentially a proof-of-concept for their own product.
• LEGaTO
  Their largest funding award (EUR 177,750) and their entry into a multi-country RIA consortium tackling low-energy heterogeneous computing — demonstrating technical depth well beyond typical AI consultancies.