Cyera says Claude reaches 88% weekly usage across its 1,500 employees
A new Anthropic case study describes the permissions work, employee training and human review behind Cyera’s company-reported adoption and time savings.
Loading page…
A new Anthropic case study describes the permissions work, employee training and human review behind Cyera’s company-reported adoption and time savings.
Listen to this story
Cyera’s rollout hinged on making company systems safe and understandable to use: it connected Claude Cowork to business tools, audited permissions, and documented Snowflake data with departmental definitions. The company says it reached all 1,500 employees in 17 days, with 88% weekly active usage; support workflows draft answers for human review, while legal and operations agents automate routine checks. The adoption figures and savings are from an Anthropic-published customer case study, not an independent evaluation.
Cowork connected to 40 tools, including Slack, Salesforce, Asana, Google Drive, and Notion.
Cyera documented 40 Snowflake tables in a departmental semantic layer and adjusted permissions after mapping access to production data.
A scheduled support-channel task checks messages roughly every 30 minutes and drafts replies for an employee to review.
Claude Enterprise has reached 88% weekly active usage across roughly 1,500 employees at cybersecurity company Cyera, according to a new Anthropic case study. The account describes a company-wide deployment built around connected business tools, with AI handling first drafts and routine checks while employees retain responsibility for reviewing the work.
The adoption figures and workflow savings are Cyera’s results published by its AI supplier, not an independent evaluation. The account nevertheless offers a detailed look at the work surrounding the rollout: access permissions, usable company data and sustained employee training.
Claude Code had already spread through much of Cyera’s research and development organization. Bringing other departments along was harder. Principal technologist Joe Tustin describes teaching nontechnical colleagues to use the terminal—the text interface for running computer commands—but finding that many remained uncomfortable executing commands or connecting to Snowflake, the company’s data platform.
The team considered building its own interface around Claude Code before choosing Cowork, Claude’s tool for nontechnical task work. Cyera connected Cowork to 40 tools, including Slack, Salesforce, Asana, Google Drive and Notion. Its aim was a general-purpose agent employees could use to address their own workflows, rather than an engineering-dependent setup.
That breadth required preparation. Before rollout, Cyera mapped which systems Claude would connect to, which held production data and how activity would be monitored. It used its own security platform to identify who could reach particular Snowflake tables, then adjusted permissions where needed. Agents, the case study notes, can use whatever their permissions allow.
Access alone did not make the data usable. Data engineers built a semantic layer—a shared set of definitions for interpreting company data—covering 40 documented tables organized by department. They paired it with instructions explaining Cyera’s terminology and conventions, giving employees who had never queried Snowflake a way into self-service analytics.
After a brief pilot with leaders from different departments, Cyera says it rolled out Cowork company-wide in 17 days. Support included a full day of livestreamed demonstrations, twice-weekly office hours and 20 department-specific sessions. The AI team also worked with system owners to establish what employees needed and how Claude should access it.
The internal Ask Claude support channel illustrates that division of labor. A scheduled task checks for new messages about every 30 minutes, compares them with knowledge bases and previous answers, and drafts replies for the owner to review. The automation handles the repetitive preparation rather than sending answers without oversight.
Cyera’s VP of AI, Steve Klementowski, describes a change in the analytics team’s work as employees begin serving themselves: less time producing reports, more time improving the data and context behind queries. With access established, he says the next focus is continued training and helping teams pursue their most strategic projects.
Loading discussion...
Join the conversation
Explain where you would draw the line on access.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.