Cymphony

What if you could cut AI-driven data exposure by 74% in your first 30 days, without deploying a single agent? Security teams using Cymphony did. See how Cymphony works.

“The workforce is no longer just human." Oindrilla (Dri) Choudhury, Human Risk and Incident Management Lead at Syngenta Group, said this about what shifted for her team once AI tools and agents started working next to their employees. Her team began with identity hygiene. They mapped the digital identities across Syngenta's environment and what each one could reach. With identity, data and behavior in one view, Dri's team could study how people and machines interact inside the company. She describes workforce risk as something that used to feel intangible and is now "measurable, actionable, and most importantly achievable." Her advice to security teams starting out is plain: "Fix your basics and then just build from there, 'cause it becomes much easier." Syngenta's environment is unusual, and Dri wanted a team that would understand it and build with her. Thank you for letting us be that team, Dri. 🙏

Muse is already being connected to your company’s most valuable assets. And it’s happening fast. Meta's Muse launched less than a month ago. It has already passed 5 million downloads, making it the fastest-growing AI agents in the US. Last week, at one of the companies we work with, Cymphony surfaced employees who had connected Muse to some of the company’s most sensitive data repositories. The interesting part? We didn’t have to teach Cymphony what Muse was. No new integration. No new detection rule. No waiting for the security team to add another AI tool to a list. Cymphony’s Workforce Security Graph automatically identified a new AI agent entering the environment, mapped who was using it and what it could access, and surfaced the risk. This is what securing AI adoption has to look like. New agents will emerge faster than security teams can keep up with them manually. Some will be sanctioned. Many won’t. Your security platform needs to understand the workforce as it changes, not just the tools it already knows. That’s the power of the Workforce Security Graph.

𝝞𝗳 60% 𝝾𝗳 𝘆𝝾𝘂𝗿 𝝾𝗿𝗴𝗮𝗻𝗶𝘇𝗮𝘁𝗶𝝾𝗻’𝘀 𝗱𝗮𝘁𝗮 𝗶𝘀 𝘀𝗲𝗻𝘀𝗶𝘁𝗶𝘃𝗲, “𝗳𝗶𝗻𝗱 𝘁𝗵𝗲 𝘀𝗲𝗻𝘀𝗶𝘁𝗶𝘃𝗲 𝗱𝗮𝘁𝗮” 𝗶𝘀𝗻’𝘁 𝗺𝘂𝗰𝗵 𝝾𝗳 𝗮 𝗿𝗶𝘀𝗸 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝘆. In a recent POC we found tens of thousands of files shared publicly. The issue was that nobody could say which ones were actually a problem. The team's first instinct was to ask if the files belonged to marketing - because being publicly shared doesn't automatically make them risky. What actually matters is who owns the file, whether anyone's still using it, and if it makes sense for the business to have it sitting there in the first place. Scanning and labeling every file in an organization can take years, and a lot of teams go through that whole process without getting any closer to reducing risk. Connecting the file to its owner, usage, and the business context around it is what lets a team skip past cataloging and go straight to fixing what actually matters.

Tools suck. Buy the wrong tool and you make whatever problem you have a lot worse. “Data” or “Identity” have been nagging problems, and my tools create more findings and alerts, putting more pressure on my team AI broke the dam and I needed to start solving. I wanted to map my organization, understand the risk, and focus on reducing risk and doing work rather than generating more findings. What Cymphony changed for us is the starting point: one connected view of who and what has access, what they're doing with it, and what deserves attention first. This is what allows us to drive action, FAST. Congratulations to Shy Dekel Elinor Sadka, Idan Berkovits Erez (Edi) Gotlieb and the team, we want to see much more of these guys.

Every so often, a cybersecurity category has to be rebuilt because the thing it protects changed shape. It's happening again, and faster this time. AI agents are being added to the workforce before security teams have a clear view of what's happening. The critical unanswered question is what agents, and on whose behalf, have access to what data and systems. The legacy identity and data security tooling was built for a static, siloed world where you could put human employees in RBAC boxes. But AI agents shape-shift all the time; they download new skills, gain access to new systems etc. The previously siloed disciplines of identity and data security now need to become one. Shy Dekel, Idan Berkovits, and Erez (Edi) Gotlieb, saw that shift early and built for it. Every security leader we spoke with recognized the problem in the first two minutes. We at Sequoia Capital are excited to co-lead Cymphony's $30M round. The best signal during diligence is what customers say. Congratulations Shy Dekel, Idan Berkovits, Erez (Edi) Gotlieb, and team!