Both MMT and QT21 projects directly involved building and improving automated translation systems, with MMT targeting a language-independent commercial MT service.
TAUS BV
Dutch language technology SME specializing in machine translation quality, open-source MT infrastructure, and translation industry data intelligence.
Their core work
TAUS BV is a Dutch technology SME and industry intelligence company operating at the intersection of machine translation, language data, and translation automation. Their H2020 participation focused on building open-source commercial machine translation infrastructure (MMT project) and improving the quality of automated translation systems (QT21 project). TAUS has an established role in the language technology sector as a data broker and standards body, connecting language service providers, technology vendors, and enterprises around shared translation data and MT benchmarks. Their work is applied and commercially oriented — they build and validate systems that make machine translation usable in real business contexts.
What they specialise in
QT21 (Quality Translation 21) was specifically focused on advancing the quality of machine translation output and establishing evaluation standards.
MMT was built on a new open-source foundation, indicating TAUS contributes to shared, reusable MT infrastructure rather than purely proprietary systems.
TAUS's participation in both major EU MT projects of the 2015 era reflects their role as a data and standards authority for the translation industry.
How they've shifted over time
Both H2020 projects ran from 2015, making a meaningful early-versus-late evolution analysis impossible from this dataset alone — there is no keyword or thematic shift to observe across phases. What can be said is that TAUS entered H2020 with a clear, focused identity: machine translation quality and commercial MT infrastructure, not a broad digital services profile. Their absence from later H2020 calls (2016–2020) may indicate that their EU project activity was selective and strategic rather than continuous, or that their primary work moved outside the Horizon funding framework.
Based on H2020 data alone, TAUS shows a stable, narrow specialization in machine translation rather than a visible pivot — anyone planning a future collaboration should verify whether their current portfolio has expanded into adjacent areas such as neural MT, multilingual NLP, or AI-assisted localization.
How they like to work
TAUS participated in both projects as a partner, never as coordinator, suggesting they prefer joining technically strong consortia where they can contribute specialized data and industry expertise without carrying full project management responsibility. With 14 unique partners across 9 countries from just 2 projects, their network is remarkably broad relative to their project volume, indicating they operate in well-connected European research and industry circles. This makes them a likely high-value partner for language technology consortia that need an industry bridge between research outputs and commercial adoption.
TAUS reached 14 unique consortium partners across 9 countries through only 2 projects — an unusually dense network for such a small project portfolio, reflecting their embedded position in the European language technology ecosystem. Their collaborations span multiple EU countries, consistent with the naturally multilingual nature of translation technology work.
What sets them apart
TAUS occupies a rare position as an SME that functions simultaneously as a data company, industry think tank, and technology partner — most organizations in the language technology space are either pure research groups or pure vendors. Their involvement in both MMT and QT21 shows they are trusted by academic and industrial consortia alike, which makes them a credible bridge partner when a project needs to demonstrate commercial relevance for translation or multilingual AI outputs. For a consortium coordinator, TAUS brings not just technical contribution but industry legitimacy and access to the broader translation and localization market.
Highlights from their portfolio
- MMTThe largest project by funding (EUR 630,000) and the most commercially ambitious — building a language-independent open-source MT service positions this as infrastructure-level work with direct market application.
- QT21Quality Translation 21 addressed one of the core unsolved problems in MT — output quality evaluation — making it scientifically significant and practically relevant to any organization deploying automated translation.