From LinTO to Whisper: six years of French innovation in audio transcription | Linagora

From LinTO to Whisper: six years of French innovation in audio transcription

From LinTO to Whisper: six years of French innovation in audio transcription

FOSDEM2026, the largest European meeting of free software and open source, took place a few days ago in Brussels. We shared, exchanged, spoke, debated the advances of our ecosystem… And we also had a lot of fun.
Because that’s what a community like ours is:

  • intense discussions about the future of free digital,
  • meetings with people coming from the four corners of Europe,
  • and laughs around a box of Belgian fries

Benjamin Bellamy, our Business Development Manager AI Solutions & Community Management, presented two talks:

1. Amplifying our voices without compromise
How decentralized and Open Source technologies like Fédiverse enable content creators to regain control of their audience and their data, without depending on third‑party platforms.

2. Rethinking audio publishing
A user‑centred approach focused on autonomy and respect for privacy.

These interventions showed that open protocols such as RSS and ActivityPub are not only for publishing content: they allow discovery of content, the creation of communities, and all of this without going through the algorithms of centralized American platforms. Accessibility and discoverability of audio content are at the heart of this approach.

A subject that is close to our hearts. For 6 years, LinTO has been providing these capabilities to French‑language content through automatic transcription, well before solutions like OpenAI’s Whisper arrived. This sovereign approach lets creators make their productions accessible to deaf and hard‑of‑hearing people, while improving their discoverability, without relying on foreign technologies.

Radios and media, creators, developers: how do you imagine a future where audio is sovereign, accessible and decentralized?


FOSDEM 2026 - Benjamin Bellamy &  Andrzej Neugebauer