Tellia raises €4.3m pre-seed to let farm teams record data by talking
Revent leads a €4.3m pre-seed for Tellia, whose voice AI turns spoken field notes into structured records inside existing farm software.
Tellia, a voice AI company for agriculture founded in 2024, announced a €4.3 million pre-seed round on 8 September 2026. The round is reported as $5 million in dollar terms. Revent led it, with Grey Silo Ventures, the corporate venture arm of Cereal Docks Group, Jeriko, operated by Techmind, and Fund F participating. Coline Labadie de Faÿ is chief executive and co-founder. EU-Startups describes the company as Paris-based; Startups Magazine and iGrow News report teams in both Paris and San Francisco.
The product replaces typing with speaking. Farmers, agronomists and field teams record what happened through a phone call, a voice note, a text message, WhatsApp or email, and Tellia converts that unstructured input into records tied to specific fields, crops and crews. There is no fixed template to fill in, and the system answers questions back against the data it has captured.
The distribution model is embedded rather than direct. Partners integrate Tellia through an API, so the voice layer sits inside the agricultural software a customer already runs and returns structured records and recommendations there. Proceeds go to advancing the voice AI itself, building an agentic layer that connects field operations to existing AgTech platforms, and expanding the team in the United States and Europe.
“The Ag industry keeps facing the same challenge: the people who spend their days in the field spend their evenings at a keyboard,” said Labadie de Faÿ. Otto Birnbaum, general partner at Revent, said: “We believe that the voice-first approach will lead the agricultural sector into the new agentic data age.” EU-Startups places the round alongside other European agrifood raises this year, including Rainbow Crops at €9.7 million, Biographica at €8 million and UBEES at €8 million, a cluster that puts European ag AI cheques at this stage firmly in single-digit millions.
Why it matters for AgTech founders: data capture has been the unfixed floor of the ag software stack for twenty years, and every farm management system built on top of it quietly assumes somebody types. What is worth studying here is not the model but the go-to-market. Tellia is not asking a farm to adopt a fifth application; it is selling the input layer into systems of record that already have the account. That reframes two decisions. If you run an FMS or an agronomy platform, voice input has just become a buy-or-build question with a closing window, and the honest answer for most teams is buy. If you build point solutions, the defensible asset is no longer the interface but the ontology underneath it, the field, crop, operation and crew objects that spoken input has to resolve against, because whoever owns that resolution owns the data. The caution is the stage. A €4.3 million pre-seed is priced on a team and a thesis, not on deployed acres, and nothing published so far shows whether operators keep talking to it in week twelve of a season rather than week three. Retention across a full season, not demo accuracy, is the number to ask for.
Tellia raises €4.3m pre-seed to let farm teams record data by talking
Revent leads the round for the voice AI company founded in 2024, with Grey Silo Ventures, Jeriko and Fund F participating. The product turns spoken field updates into structured records inside other people's software.
Tellia, a voice AI company for agriculture founded in 2024, announced a €4.3 million pre-seed round on 8 September 2026. The round is reported as $5 million in dollar terms. Revent led it, with Grey Silo Ventures, the corporate venture arm of Cereal Docks Group, Jeriko, operated by Techmind, and Fund F participating. Coline Labadie de Faÿ is chief executive and co-founder. EU-Startups describes the company as Paris-based; Startups Magazine and iGrow News report teams in both Paris and San Francisco.
The product replaces typing with speaking. Farmers, agronomists and field teams record what happened through a phone call, a voice note, a text message, WhatsApp or email, and Tellia converts that unstructured input into records tied to specific fields, crops and crews. There is no fixed template to fill in, and the system answers questions back against the data it has captured.
The distribution model is embedded rather than direct. Partners integrate Tellia through an API, so the voice layer sits inside the agricultural software a customer already runs and returns structured records and recommendations there. Proceeds go to advancing the voice AI itself, building an agentic layer that connects field operations to existing AgTech platforms, and expanding the team in the United States and Europe.
“The Ag industry keeps facing the same challenge: the people who spend their days in the field spend their evenings at a keyboard,” said Labadie de Faÿ. Otto Birnbaum, general partner at Revent, said: “We believe that the voice-first approach will lead the agricultural sector into the new agentic data age.” EU-Startups places the round alongside other European agrifood raises this year, including Rainbow Crops at €9.7 million, Biographica at €8 million and UBEES at €8 million, a cluster that puts European ag AI cheques at this stage firmly in single-digit millions.
Why it matters for AgTech founders: data capture has been the unfixed floor of the ag software stack for twenty years, and every farm management system built on top of it quietly assumes somebody types. What is worth studying here is not the model but the go-to-market. Tellia is not asking a farm to adopt a fifth application; it is selling the input layer into systems of record that already have the account. That reframes two decisions. If you run an FMS or an agronomy platform, voice input has just become a buy-or-build question with a closing window, and the honest answer for most teams is buy. If you build point solutions, the defensible asset is no longer the interface but the ontology underneath it, the field, crop, operation and crew objects that spoken input has to resolve against, because whoever owns that resolution owns the data. The caution is the stage. A €4.3 million pre-seed is priced on a team and a thesis, not on deployed acres, and nothing published so far shows whether operators keep talking to it in week twelve of a season rather than week three. Retention across a full season, not demo accuracy, is the number to ask for.
