PHONEXIA SRO

Their core work

Phonexia is a Czech speech technology company specializing in voice biometrics, speech analytics, and natural language processing. They build software that extracts actionable intelligence from audio data — whether that means analyzing customer calls in contact centers or processing intercepted communications for law enforcement. Their core technology covers speaker identification, language recognition, keyword spotting, and emotion detection across multiple languages. As an SME based in Brno, they operate at the intersection of commercial speech analytics and security-oriented audio intelligence.

What they specialise in

How they've shifted over time

Phonexia entered H2020 with a clear commercial focus: big data speech analytics for contact centers and business intelligence (BISON, MixedEmotions, both 2015-2017). From 2019 onward, their work shifted decisively toward security applications and deeper research — criminal network analysis, counter-terrorism audio processing (ROXANNE), and fundamental speech science including explainability and low-resource languages (ESPERANTO). The trajectory shows a company moving from pure business analytics toward dual-use speech technology that serves both commercial and law enforcement needs, while investing in the scientific foundations of their core technology.

Phonexia is moving toward explainable, trustworthy speech AI that can operate in low-resource languages — positioning them for security, humanitarian, and multilingual market applications.

How they like to work

Phonexia primarily joins consortia as a specialist partner (3 of 4 projects), contributing their speech technology to larger platforms. They coordinated one project (BISON), demonstrating they can lead when the topic aligns tightly with their core product. With 56 unique partners across 26 countries, they are well-connected and comfortable working in large, diverse European consortia rather than sticking to a narrow set of repeat collaborators.

Phonexia has built a broad European network of 56 unique partners across 26 countries through just 4 projects, indicating participation in large multi-partner consortia. Their reach spans well beyond Central Europe, covering a wide geographic spread typical of ICT and security research collaborations.

What sets them apart

Phonexia brings production-grade speech technology into research consortia — they are not a lab, but a company with deployable products in voice biometrics and audio analytics. This makes them a rare bridge between academic speech processing research and real-world deployment in both commercial (contact centers) and security (law enforcement) settings. For consortium builders, they offer a credible SME that can turn research prototypes into working software, with proven experience across both civilian and security domains.

Highlights from their portfolio

• BISON
  Their only coordinator role, with the largest single grant (EUR 385K), focused squarely on their core commercial product — big speech data analytics for contact centers.
• ROXANNE
  Marked their strategic pivot into security, applying speech analytics to real-time criminal network analysis and counter-terrorism — a high-impact domain with significant funding (EUR 345K).
• ESPERANTO
  A research mobility project (MSCA-RISE) signaling investment in fundamental speech science — explainability, low-resource languages — building long-term scientific credibility.