Contributed UML, MARTE, and TOSCA modeling expertise in the DICE project focused on quality enhancement for data-intensive cloud applications.
NETFECTIVE TECHNOLOGY SA
French software SME specialising in cloud quality engineering and AI-driven voice interfaces with privacy-by-design.
Their core work
Netfective Technology is a French software engineering SME that builds tooling and methods for quality-assured, scalable software systems. In their first EU project, they contributed to model-driven engineering approaches for data-intensive cloud applications, applying standards like UML, MARTE, and TOSCA to improve reliability and safety. In their second project, they shifted toward AI-powered voice interaction, bringing privacy-by-design principles and deep learning into multilingual speech services. They appear to operate as a specialist software partner — bringing technical depth in a specific area rather than broad system integration.
What they specialise in
Applied weakly supervised learning and deep learning to multilingual spoken interaction in the COMPRISE project.
COMPRISE explicitly addressed privacy-driven voice services, with privacy-by-design listed as a core keyword alongside personalisation.
DICE centred on iterative quality enhancements for cloud applications, with quality, reliability, efficiency, and safety all named as defining themes.
How they've shifted over time
Between 2015 and 2018, Netfective's work was grounded in formal software engineering — model-driven methods (UML, MARTE), cloud orchestration standards (TOSCA), and quality assurance for big-data applications. Their second project, starting in 2018, marks a clear pivot toward applied AI: spoken interaction, deep learning, weakly supervised learning, and personalisation replaced the earlier engineering-tooling vocabulary entirely. The trajectory suggests a deliberate move from software quality infrastructure toward user-facing AI features, likely tracking commercial demand in the late 2010s.
Netfective appears to be repositioning from back-end software engineering tooling toward AI-powered application layers — a direction that aligns with growing demand for privacy-compliant, personalised voice and language services.
How they like to work
Netfective has participated in both projects strictly as a partner, never taking on a coordinator role, suggesting they prefer contributing specialist expertise within larger consortia rather than managing projects. With 16 unique partners across 9 countries over just two projects, their consortia have been mid-to-large in size. There is no evidence of repeat partnerships, indicating they integrate into diverse teams rather than maintaining a fixed collaboration circle.
Netfective has collaborated with 16 distinct organisations spread across 9 countries, a notably broad reach for a company with only two EU projects. Their network spans multiple European nations, consistent with participation in geographically diverse RIA consortia.
What sets them apart
Netfective occupies an unusual position as a small French software company that bridges formal software engineering rigour (model-driven methods, quality standards) with applied AI — two domains that rarely combine in one SME. For consortium builders, they offer technical depth in areas — TOSCA orchestration, MARTE profiling, privacy-by-design — that are hard to source from generalist IT firms. Their SME status and track record in competitive RIA projects also makes them an attractive partner for projects seeking to demonstrate industrial relevance.
Highlights from their portfolio
- DICELargest grant received (€567,875) and grounded in technically demanding model-driven cloud engineering, covering quality, safety, and big-data orchestration — a rare combination for an SME.
- COMPRISEMarks a significant thematic shift into AI and privacy, demonstrating Netfective's ability to enter a new technical domain mid-career and contribute deep learning expertise to a multilingual voice project.