Deals · Enterprise AI
Mio raises €1.9M pre-seed to build an AI colleague that lives inside Slack
A Paris startup building an AI 'colleague' that lives inside Slack — learning a company's own documents, messages, and tools, and proactively executing recurring tasks such as meeting recaps, CRM updates, and project briefs without requiring explicit prompts.
“We don't think the future of AI at work is another app people have to open. The best colleague already understands what's happening, knows how your company works, and steps in before you have to ask.”
Mio, a Paris-based startup founded in 2025 by Arthaud Mesnard (CEO), has raised €1.9 million ($2.2 million) in a pre-seed round co-led by Fabric.vc and Topology.vc. The company simultaneously emerged from stealth. The capital will fund engineering team growth, product development, and customer acquisition in Europe and the United States.
Mio builds what it calls an AI "colleague" embedded natively inside Slack. Unlike an AI assistant that requires a user to prompt it, Mio is designed to act proactively. It learns a company's own context — past messages, documents, connected tools — and takes over recurring tasks without being explicitly asked: generating meeting summaries, updating CRM records, writing project briefs, and synthesising cross-channel context into actionable briefs. The product integrates with Google Workspace, Notion, Linear, HubSpot, and GitHub. Early customers report average savings of 8.2 hours per week per user.
The context problem in workplace AI
Most enterprise AI tools face the same limitation: they are general-purpose systems applied to specific company contexts via document upload or custom prompt engineering. The result is a capable assistant that doesn't actually know how this company works — who the clients are, what the recurring reports look like, what the shorthand terms mean. Every session starts near-zero.
Mio's architecture inverts this. By living inside Slack — where most knowledge-worker communication at software-forward companies happens — the assistant accumulates company-specific context continuously and passively. It is not trained on a company's data in the model-training sense; it reads and acts on live context. The distinction matters for both privacy and performance: the model doesn't need to be fine-tuned to be useful, but it does need persistent access to the communication layer.
"We don't think the future of AI at work is another app people have to open," Mesnard said. "The best colleague already understands what's happening, knows how your company works, and steps in before you have to ask."
A pre-seed with structural competition
At €1.9 million, this is early capital buying product time, not market proof. The field Mio is entering — AI agents for team productivity — is active and well-funded. Slack itself has been building AI features natively. Microsoft Copilot occupies a structurally similar position in Microsoft 365. Notion, Linear, and HubSpot — all of which Mio integrates with — have their own AI layers.
The Mio bet is that a Slack-native agent with persistent, proactive behaviour is a meaningfully different product from an AI feature bolted onto an existing tool. Whether that differentiation survives the incumbents' roadmaps is the core risk of the investment. Fabric.vc and Topology.vc, both European early-stage funds with software-as-a-service backgrounds, are backing the thesis that there is a durable position at the cross-tool coordination layer before the large platforms absorb it.
Sources
Threaded to this story
Enterprise AI ·
Mendo raises €12M Series A to close enterprise AI's adoption gap
€12M · Series A
Enterprise AI ·
Skalar raises €12M to build the AI accounting firm that replaced human firms — not software selling to them
€12M
Enterprise AI ·
Sherpa raises €2M Pre-Seed to manage external workers and AI agents through one platform
€2M · Seed
Every European round, in your inbox by 8am.
The day's seed and Series A rounds across France and Europe — threaded, sourced, and read in two minutes. Free.