Archestra AI raises $3.3m to secure enterprise AI agents

London-based platform Archestra AI has exited stealth mode and secured $3.3 million in pre-seed investment to address the growing challenge of safely connecting AI agents to company data.

The round was led by Concept Ventures, with participation from Zero Prime Ventures, Celero Ventures, RTP Global and Aloniq. It also attracted support from industry figures including Max Hauser, Managing Director and Partner at BCG; Maxim Konovalov, Nginx co-founder; Konstantin Vinogradov, General Partner at Runa Capital; Stephen Whitworth, CEO of incident.io; and Luke Harries from Elevenlabs.

Preventing risks

Archestra's founders, CEO Matvey Kukuy and CTO Ildar Iskhakov, previously led incident management platform Amixr, which was sold to Grafana in 2021. Kukuy subsequently co-founded AIOps platform Keep, later acquired by Elastic. They are joined by Founding Engineer Joey Orlando, who previously worked at Grafana.

The company's platform is aimed at allowing organisations to leverage the capabilities of AI agents while ensuring security and compliance remain intact. It does so by introducing security frameworks for linking AI agents - such as those utilising Large Language Models (LLMs) like ChatGPT - with sensitive company data and internal systems.

With enterprises increasingly seeking to unlock the productivity benefits of AI, the risks of unfettered AI access have become more apparent. Potential scenarios include unauthorised access and sharing of sensitive information, such as salary data, or accidental deletion of critical business information.

MCP approach

Archestra's solution builds on the recently introduced Model Context Protocol (MCP), a framework unveiled by AI company Anthropic. MCP enables organisations to connect LLMs with a wide variety of systems and data sources, from HR platforms to messaging tools like Slack, supporting a range of tasks such as drafting emails, editing documents, and accessing internal databases.

The challenge, according to Archestra, lies in enterprise adoption, where security and governance barriers remain significant.

"MCP is unlocking a new frontier for AI agents. But, right now, MCP is completely unsuitable for the enterprise. We're building the security-first solution that will change this. Our open-source platform means anyone, whether you're an engineer or a HR rep, can safely and impactfully integrate AI agents into their workflow - driving efficiency and impact at scale - without increasing risks. The speed at which we're moving reflects how passionate we are about this solution and the impact we think it can have in the market."

This statement from Matvey Kukuy reflects the company's view that there is an urgent need for robust security when it comes to enterprise AI adoption. The open-source nature of the platform is designed to make its tools widely accessible and adaptable for different organisational roles and functions.

Market opportunity

Archestra's product targets the integration-platform-as-a-service (iPaaS) market, which industry forecasts suggest will exceed $17 billion in revenue by 2028. As autonomous AI agents become increasingly central to business automation and operations, the demand for secure and scalable infrastructure is expected to rise.

Backing this vision, Ariel Rahamim, Principal at Concept Ventures, commented:

"Just as APIs became the foundational building blocks for internet infrastructure, MCPs are emerging as the connective tissue for improving the context layer of AI tools within enterprise. Archestra is building the infrastructure layer this ecosystem needs. At Concept, we leaned into the teams' technical depth, open-source DNA, and executional intensity to grow into a leader in the category. The team is attacking a mission-critical problem they've lived firsthand, and we're confident they'll define how enterprises deploy and govern AI context at scale."

The oversubscribed nature of the funding round, which closed in less than two weeks, reflects significant market interest in solutions that enable safer enterprise adoption of AI-driven workflow automation.