HOPSWORKS AB

Their core work

Hopsworks AB is a Swedish technology SME that builds scalable data platforms and machine learning infrastructure for processing and analyzing large datasets. In EU-funded projects, they provide the underlying data engineering layer — feature stores, AI pipelines, and platform-as-a-service (PaaS/IaaS) components — that allows consortium partners to run deep learning and explainable AI workloads on massive datasets such as Copernicus Earth observation data and human health exposome records. Their core contribution is turning raw big data into AI-ready infrastructure that domain scientists can actually use.

What they specialise in

How they've shifted over time

Hopsworks entered H2020 in 2019 with ExtremeEarth, focused squarely on big data infrastructure for Copernicus satellite analytics. By 2020-2021, they expanded into health data (HEAP, their largest project) and explainable AI (DeepCube), showing a deliberate move from pure data engineering toward AI transparency and cross-domain data platforms. The shift from generic big data processing toward FAIR data principles, explainable AI, and domain-specific applications (health exposome, climate) suggests a maturing platform company seeking higher-value positioning.

Hopsworks is moving from backend data infrastructure toward AI-ready platforms with explainability and FAIR compliance — positioning for the growing demand for trustworthy, regulation-ready AI systems.

How they like to work

Hopsworks operates exclusively as a participant, never as coordinator, which is typical for a technology SME that supplies platform components rather than driving research agendas. With 29 unique partners across 13 countries from just 3 projects, they work in large, diverse consortia — likely providing shared infrastructure that many partners depend on. This makes them a reliable technology provider that integrates well into multi-partner setups without needing to lead.

Despite only three projects, Hopsworks has built a broad European network of 29 partners across 13 countries, reflecting their role as a shared platform provider in large consortia spanning Earth observation, AI, and health research communities.

What sets them apart

Hopsworks occupies a distinctive niche as a dedicated ML data platform company participating in research consortia — they are not a university lab or a consulting firm, but a product company whose platform becomes shared infrastructure for the entire project. Their ability to serve radically different domains (satellite imagery, human health, climate) with the same underlying data and AI platform makes them unusually versatile. For consortium builders, they solve the common problem of "we have the data and the domain experts, but no scalable ML infrastructure."

Highlights from their portfolio

• HEAP
  Largest grant (EUR 1.35M) and longest project (2020-2025), representing a major investment in health exposome data infrastructure with FAIR principles and wearable sensor integration.
• DeepCube
  Focused specifically on explainable AI for Copernicus data — a timely combination of AI transparency requirements with Earth observation at scale.