Ex-DeepMind founders raise $20M for AI sales startup Airspeed

Two former Google DeepMind researchers have secured $20 million in Series A funding for their London-based startup, Airspeed, which is building what it calls an AI execution layer for revenue teams. The round was led by DN Capital, with participation from Vi Partners, Framework Venture Partners, and Atlassian Ventures, bringing Airspeed’s total funding to over $25 million.

The pitch for most AI sales tools follows a predictable pattern: more signals, better insights, and smarter forecasting. What rarely gets mentioned is that those insights often go unused. Sales representatives drown in dashboards. CRM records fall out of date. A deal that should have closed slips away because nobody acted on the intent signal that arrived days earlier. Airspeed aims to solve that problem by deploying autonomous AI agents that don’t just surface insights but act on them directly.

The agents work across calls, emails, support tickets, and CRM systems. Instead of waiting for a human to notice a flag or follow up on a lead, they update records, generate follow-ups, flag risks, and route tasks automatically. Co-founders Adam Liska and Devang Agrawal left DeepMind in 2022 to start the company from a small WeWork in London. Liska, who holds a PhD in brain and cognitive sciences and previously worked at Spotify and Facebook, frames the challenge as one of execution rather than intelligence. “You can have all the data in the world and still lose because no one executed at the right moment, in the right way,” he wrote in announcing the round.

Agrawal, a Cambridge engineering graduate who studied on a Dr Manmohan Singh Scholarship, noted that the relationship with DN Capital began not in a boardroom but at a customer event in Stockholm, where the investor witnessed firsthand the intensity of customer engagement. The company was originally called Glyphic and rebranded to Airspeed on 20 May 2026.

Airspeed reports that it has grown revenue four-fold year over year and now serves nearly 200 customers across 20 countries, including identity verification platform Persona, vector database provider Qdrant, and pricing software company Pricefx. Customers have built thousands of custom agents on the platform in the first four months of 2026, with monthly run volume nearly tripling between January and April, according to VentureBurn. The company operates across London and New York and plans to use the new funding for product development and go-to-market hiring.

The market Airspeed enters is crowded. Revenue intelligence and sales engagement tools from companies like Gong, Clari, and Outreach already command significant enterprise budgets. Airspeed bets that the next wave is not about better analytics but about agentic execution , AI that does the work rather than telling humans what work to do. The involvement of Atlassian Ventures is particularly telling. Atlassian has been building its own agent ecosystem around Confluence and Jira, and a strategic investment in an agentic go-to-market platform suggests the company sees revenue operations as adjacent to its own product roadmap.

Whether agentic AI in sales represents a genuine shift or just the latest packaging for CRM automation will depend on whether tools like Airspeed can demonstrate measurably better outcomes than the dashboards they aim to replace. The DeepMind pedigree of the founding team lends credibility to the technical bet, but the harder test is commercial: convincing revenue leaders that an agent they cannot see is doing the job better than the rep they can.