Netgem Acquires Israeli AI Localization Startup BARY

According to The Jerusalem Post, French media and technology group Netgem, which is traded on Euronext, is acquiring Israeli AI startup BARY, the developer of TheSubtil.ai, an AI-driven platform for translating, subtitling, and dubbing films and series. Under the deal Netgem will acquire BARY's technology assets and its research and development unit in Israel. The company's founders are Matthias Cohen-Scali, Elie Zarbiv, and Sasha Rebouh.

The Jerusalem Post reports that BARY says its technology reduces content localization costs by about 50%, cutting a typical 18 euro per-minute cost to about 9 euros per minute. TheSubtil.ai has translated thousands of films, series, and TV programs into more than 25 languages, with particularly strong demand in regional European languages including Catalan. The technology was also used on thousands of works shown at the Cannes Film Festival and the Venice Biennale, per the article.

"Our main advantage is the combination of an advanced AI engine with strict human oversight," said Matthias Cohen-Scali, per the Jerusalem Post. "The human element is essential to preserve cultural nuances, humor, and local expressions that technology alone cannot always convey."

Following the acquisition, Eclair - part of the Netgem Group and one of the oldest companies in the film services sector, founded in 1907 - will fully integrate TheSubtil.ai and make it a central component of its localization activities in France and international markets. Even before the acquisition, BARY demonstrated its capabilities through cooperation with the M6 Group, one of France's largest media and television groups, which supported product development, per the Jerusalem Post.

BARY is also developing the next generation of accessibility solutions, focused on subtitling systems for the hard of hearing and deaf (SDH), a field facing growing regulatory and commercial requirements from streaming platforms worldwide.

AI-first localization pipelines generally combine machine translation, automated subtitle timing, and voice or lip-sync tools with human post-editing to reach broadcast quality. The persistent trade-off is between speed and cost versus linguistic nuance: automated pipelines deliver scale and lower marginal costs while human-in-the-loop steps address cultural references, idioms, and brand voice. Regional languages - exactly those cited in the report - are a common vector where automation yields outsized ROI because human specialists are scarce and per-minute costs rise quickly. For practitioners, SDH subtitling is also a regulatory pressure point as streaming regulations tighten across European markets.