Cymphony has secured $25 million in Series A funding, a round co-led by Sequoia and SMBC Fin Atlas Beyond Fund. This brings the two-year-old startup’s total funding to $30 million, pushing its post-money valuation past the $100 million mark. Based in New York and Tel Aviv, the company is now part of a rapid consolidation of capital in the agent security sector.
This deal marks the third major funding event in this niche in just three weeks, following significant raises by AIR and Zenity. It serves as a direct follow-up to the $435 million that has poured into the space over the last five months. Investors are clearly betting that the current wave of enterprise AI adoption is creating security vulnerabilities that existing tools simply cannot see.
Cymphony’s core product is a workforce graph that attempts to unify identity, data, and activity signals. The founders—Shy Dekel, Idan Berkovits, and Edi Gotlieb—are all graduates of the Talpiot program, a pedigree that has become a shorthand for technical rigor in the cybersecurity world, similar to the origins of Wiz.
The need for this visibility is grounded in messy, real-world deployments. At one US public company, Cymphony discovered 85,000 files that were inadvertently accessible to AI tools. In another instance, an external collaborator installed an unsanctioned version of Anthropic’s Claude, which proceeded to scan thousands of sensitive files before anyone noticed. These aren’t theoretical risks; they are the operational realities of shadow AI.
Sequoia is not just backing the company; they are using Cymphony internally to manage their own exposure. This doubling down by Sequoia suggests that the firm views agent-specific security as a prerequisite for any serious enterprise AI strategy. With customers like KKR, Syngenta, and Cass Information Systems, Cymphony has already hit seven-figure ARR within its first year of sales.
Despite the hype, the broader market is struggling to move from pilot to production. According to data from IDC and Lenovo, 88% of enterprises with agent initiatives never actually ship them to production. This massive gap between experimentation and deployment is where the security industry is trying to plant its flag, arguing that without proper governance, these projects are destined to remain in limbo.
The fundamental challenge is that enterprise security was designed for humans, not autonomous agents. Agents operate with different permissions, speeds, and access patterns than traditional users. When an agent is given access to a corporate environment, it doesn’t just act as a user; it acts as a high-velocity processor of data, often bypassing the guardrails built for human employees.
While the funding numbers are impressive, the industry faces a long road ahead. Gartner predicts that more than 40% of agentic AI projects will be canceled by the end of 2027. For companies like Cymphony, the goal is to prove that their security layer is the difference between a project that gets shut down due to risk and one that actually delivers value to the enterprise.