Sentra raises $150 million to control AI agents’ access to corporate data

• Capital One Ventures led Sentra’s $150 million Series C and will gain a board seat; total funding now exceeds $250 million.
• Sentra says it has about 100 major enterprise customers and revenue in the tens of millions of dollars, without disclosing an exact figure or current valuation.
• The company is expanding an existing data-security platform into a market with overlapping tools from established vendors.

Sentra, an enterprise data-security company founded in 2021 and headquartered in New York and Tel Aviv, announced the financing on November 4. Its pitch is to limit what AI agents can reach by securing stored data and tightening access permissions before automated systems use them.

Company-reported financing: Sentra announced a $50 million Series B in April 2025 and a $150 million Series C in November 2026. Source: Sentra.

A bigger round for an existing AI-security strategy

Capital One Ventures, the banking group’s strategic investing arm, led the round, Sentra co-founder and CEO Yoav Regev confirmed in a financing report. Regev previously spent 25 years in Israel’s military Unit 8200, most recently heading its cyber center, according to an executive interview.

The immediate problem is inherited access. AI agents—software that can take actions toward goals with limited human input—can operate with permissions broader than their tasks require. Sentra seeks to reduce the sensitive information exposed through those permissions, rather than inspecting every agent request in real time.

Sentra said in its financing announcement that existing investors Bessemer Venture Partners, Zeev Ventures, Standard Investments, Munich Re Ventures and Key1 Capital also participated. A Capital One representative will join its board. A separate financing report says a small portion of the transaction provides employee liquidity through secondary sales, with most going to the company.

The round is three times the $50 million Series B that Sentra announced in April 2025. That earlier announcement already described AI data-security capabilities and scanning within customers’ environments. The latest investment finances an expansion of that strategy; cumulative funding now exceeds $250 million.

Sentra has 130 employees, including 65 in Tel Aviv, the executive interview reports. Its founding team also includes president Asaf Kochan, a former Unit 8200 commander; chief technology officer Ron Reiter, who previously managed Oracle’s Tel Aviv development center; and Yair Cohen, who held senior roles at Microsoft and Datadog.

Growth claims leave the investment price unclear

The latest announcement claims more than 300% year-over-year growth without defining the metric. The April 2025 announcement specifically claimed more than 300% revenue growth over the preceding year. Those disclosures do not establish a consistent revenue-growth series.

The executive interview reports that Sentra declined to disclose revenue or its current valuation, citing PitchBook data that put the previous round’s valuation at about $205 million. In the separate financing report, however, Sentra says it has around 100 major enterprise customers in the United States and Europe and revenue in the tens of millions of dollars. It supplies no exact amount or reporting period. Regev also says the valuation tripled from the previous round, without giving the new figure.

Those broad company claims offer a commercial baseline, but do not establish a current revenue multiple or the ownership stakes bought in this round. Regev’s stated ambition to reach $100 million in sales is a target, not reported revenue.

Protecting stored data before agents reach it

Sentra says its platform combines discovery and classification of sensitive data, access governance, detection and response, and data-loss prevention. It describes coverage across cloud services, software-as-a-service applications, data warehouses, on-premises file shares and the knowledge stores AI systems draw on.

Its announcement describes three stages:

• Before suspicious activity: Map sensitive data and the access available to people, service accounts and agents, including inherited permissions; reduce excessive privileges and remove unnecessary, high-risk data.
• During suspicious activity: Detect unusual access, revoke permissions and remove risky sharing through automated remediation, while routing high-risk issues to data owners and existing workflows.
• Afterward: Show what an agent could have reached, with business context for the datasets involved.

A description of the platform likewise details access reporting and automated remediation. These are descriptions of capabilities, rather than published measurements of their effectiveness. An account of what an agent could reach establishes potential exposure; it does not by itself prove what the agent retrieved.

Sentra says it analyzes data entirely within the customer’s environment, without transferring sensitive information to its own systems. It argues that leaving data in place avoids a separate data estate to secure and reduces residency and compliance complications. That architecture is relevant to regulated customers, but the announcement supplies neither comparative prices nor a measured cost benchmark for its claimed cost advantage.

In the announcement, Chaim Sanders, chief information security officer at Sentra customer Lyft, describes using Sentra to understand what data agents can reach before deployment. This is a named customer endorsement supplied by Sentra; it provides no measured reduction in incidents or deployment costs.

Strategic backing meets established alternatives

Capital One explains in a September 2025 account of its investment strategy that its startup investments begin with the bank’s strategic priorities. It says it has funded more than 100 startups over the preceding decade, gaining access to technology while its technical teams offer feedback, pilots and practical use cases. The Sentra board seat gives it a governance role as well as a financial stake. The announcement does not establish a customer relationship with Capital One.

Sentra’s claim to combine data and identity controls faces existing alternatives. In August 2025 guidance, Microsoft described its Entra identity and access tools as extending governed identities and minimum necessary permissions to agents. Its Purview data-security and compliance tools address oversharing; Defender adds security-posture management and runtime threat protection. Microsoft presents these as extensions of customers’ existing security investments.

Regev also identifies Cyera and Varonis as competitors in the separate financing report. The competitive question is therefore how well Sentra’s coverage, controls and costs compare with specialist rivals and existing security suites. The available descriptions establish functional overlap, without proving equivalent coverage or performance.

What the new capital must demonstrate

Sentra says the money will support broader visibility into agents and sensitive data, continuous enforcement of minimum necessary permissions, and faster containment. It also plans to expand sales, channel programs and marketing. The executive interview describes additional research and development, hiring in Israel and abroad, and European expansion focused on London.

The separate report says Sentra plans to double its workforce in Israel and the United States, pursue U.S. federal sales beginning in the second half of 2027, and potentially make an acquisition to accelerate agent-security or data-loss-prevention development.

For customers, the next decision is whether Sentra can control more of their sensitive data, remove risky permissions reliably and justify the cost alongside tools they already own. Comparable deployment costs, measured security outcomes and precisely defined financial disclosures would help test the case for this larger investment. The financing announcement supplies none of those comparisons.