Sentra Raises $150M Series C for AI Agent Data Security

Sentra wants to inspect sensitive enterprise data without taking custody of it. That architectural choice puts the security vendor's own handling of information inside the conversation about giving AI agents access, an increasingly consequential decision for companies whose data already spans several environments.

The New York and Tel Aviv cybersecurity company closed a $150M Series C led by Capital One Ventures, announced November 4, 2026. Sentra plans to broaden its data-security platform and international operations as enterprises introduce software that can retrieve information and act on it with less human supervision. Its proposition connects two questions buyers have to resolve together: where sensitive information lives, and which identities can reach it.

The financing gives that proposition a strategic investor with a direct interest in enterprise technology. It also puts a familiar administrative task, reviewing data access, much closer to the decision about whether a business can put an agent into service.

What Sentra's $150M Series C finances

CTO Ron Reiter's announcement explicitly identifies the financing as a closed Series C. Sentra's corporate release describes the same transaction as a strategic capital investment and says a Capital One representative will join its board. The company names Bessemer Venture Partners, Zeev Ventures, Standard Investments, Munich Re Ventures and Key1 Capital as existing investors, a useful distinction from a confirmed list of participants in this particular round.

Sentra previously announced a $50M Series B led by Key1 Capital on April 22, 2025. That earlier financing brought total investment above $100M. Current Globes reporting places the cumulative figure above $250M and says the current valuation was not disclosed. The public record supports the scale of the financing without supplying an exact valuation or a detailed allocation of proceeds.

Capital One Ventures describes its investment mandate around the strategic priorities of its parent company, including data, technology and financial services. That makes data access a commercially relevant subject for the investor. The financing itself establishes an investment relationship; it supplies no basis for assuming that Capital One has purchased or deployed Sentra's software. The distinction keeps the analysis anchored in the transaction actually announced.

How data access becomes an AI deployment decision

A company can give an employee access to information for a legitimate assignment and leave that permission in place long after the circumstances change. An agent operating through an authorized identity can bring that accumulated access into a new workflow. The security question then includes the scope of the identity's permissions alongside the task the business intended the agent to perform.

In his account of Sentra's AI-security strategy, CEO Yoav Regev describes agents inheriting broad permissions and reaching sensitive information unrelated to their assignments. Sentra's response is to discover and classify data, map access across human and non-human identities, and reduce permissions that create unnecessary exposure. These are the company's stated capabilities and strategy, rather than an independent certification of their effectiveness in every environment.

The operating consequence reaches beyond the security department. A data owner may understand why a file exists while an application team understands the agent's intended work. Bringing the two together makes permission review part of implementation. A useful approval decision needs enough context to decide whether access is necessary, who can change it, and how that change affects legitimate work.

That is where a data-security supplier can become involved earlier in an enterprise AI purchase. The business is deciding what to automate while security is establishing the information the automation will be allowed to use. Both decisions shape the eventual deployment. Related DevCuration coverage of Infisical’s financing for AI agent access and Mintive’s enterprise agent governance round provides adjacent context for that work.

Why the customer's environment matters

Reiter says Sentra's scanning runs within the customer's environment and transmits metadata while leaving sensitive data there. That design gives procurement and security teams a specific architecture to examine: the place where information is processed, the permissions granted to the supplier, and the information transmitted out. Keeping data in place can address part of the handling question while leaving those other details worthy of review.

The product spans several kinds of work. Sentra describes discovery and classification, identity and access governance, detection and response, and prevention of sensitive-data loss. The connection among those functions is important to understanding the offer. A classification result becomes more useful when the company can relate it to an identity with excessive access and a workflow capable of correcting that access.

That creates a substantive buying comparison across the existing security stack. Buyers can examine coverage of their actual data stores, the quality of the classifications, the ability to explain inherited permissions, and the operational cost of making changes. An attractive demonstration still has to fit the customer's environment and decision rights. That evaluation also needs a clear owner for changing the access a tool finds excessive.

For Sentra, the architecture is also part of the trust relationship it sells. The supplier asking to examine sensitive information has to make its own access understandable to the people responsible for that information. That obligation continues as the platform reaches more systems.

The founders' next operating assignment

Sentra's current leadership page lists all four founders: Yoav Regev as CEO, Asaf Kochan as President, Ron Reiter as CTO and Yair Cohen as CPO. The company was founded in 2021. Its development from cloud-data visibility toward agent-related access controls extends an existing data-security business into another enterprise purchasing conversation.

The history matters because the data problem preceded the current agent rollout. In its early investment account, Bessemer described the attraction of finding sensitive assets customers had not known were present and simplifying deployment. More autonomous consumers of that information give the discovery work another operational use: helping a company decide what an agent should be able to reach before it begins working.

The new funding is intended to support product development and a broader commercial presence. Regev describes field and channel expansion and growth in Europe, while Reiter discusses scaling engineering and product work and deepening the company's North American and European presence. These are plans, with delivery still ahead of the team.

As those teams expand, Sentra's next customer conversation will often begin with an agent someone already wants to deploy. The people authorizing that rollout will need an intelligible account of its access, including the old permissions it inherits and the sensitive records those permissions now put within reach.