
Companies using AI agents usually pay for every interaction. The more they use an agent, the higher their bill gets. Remanence has raised $30 million in seed funding to address this recurring cost. The Paris-based startup helps businesses train and own their own models, shifting them away from renting third-party services.
Who Invested in the Round
EQT Ventures and FirstMark co-led the investment. Other participants include Zero Prime Ventures, Emblem, Drysdale Ventures, Global Founders Capital, and Yellow. The round also drew support from notable figures like Olivier Pomel of Datadog, Philippe Corrot of Mirakl, Romain Niccoli of Pigment, and Mehdi Ghissassi of AI71.
The Core Value Proposition
The AI agents most companies use today run on rented models. The more they use the AI agent, the higher their bill becomes. In the meantime, the data created by those agents sits unused. Remanence considers that pile an asset. Businesses can train their own models on their own workflow and own the model, Remanence says. Its mission is to deliver high performance, low latency, and low cost.
The platform integrates with a business’s current AI systems, capturing execution traces and converting them into customized training settings. It then employs reinforcement learning to boost open-weight models, making them less expensive to run than closed models. The more agents work, the smarter the company’s own model becomes, while the data remains relevant rather than stored in isolation.
Read Also: Plural raises €1 billion to fund European frontier tech
The founders bring deep AI experience. All three previously worked at the French AI company H Company. Cedoz is also a former Google DeepMind employee, while Hamza Benchekroun is a former executive at InstaDeep. The trio is completed by Axel Moyal.
Strategic Plans and Future Outlook
The funding will go toward expanding computing infrastructure, hiring additional engineers, and opening offices in London and San Francisco, complementing its existing base in Paris. Competitors in this space are also raising significant capital; for instance, River AI secured $1.1 billion in August to help companies train their own models.
There are no officially named customers or pilots yet.
Leave a Reply