Daylight Raises $80M as Engineers Govern AI Agents

In New York on November 10, 2026, Daylight announced an $80 million Series B led by Growth Equity at Goldman Sachs Alternatives, bringing its reported total funding to $120 million. New investor Okta Ventures joined existing backers Bain Capital, Craft Ventures, Vintage Partners and Maple Capital. The company plans to expand its U.S. presence and security engineering team.

The financing backs a managed security service in which AI agents carry out routine work while human engineers build and govern them. Engineers retain decisions requiring judgment. That division of responsibility is central to what customers buy: automation handles recurring investigations and response work, but people remain responsible for how the agents operate and for complex calls.

The round will fund expansion and hiring

Daylight’s announced financing supports expansion of its U.S. presence and security engineering team.

Daylight intends to grow its U.S. operation, enter markets across Europe, the Middle East, Africa and Asia-Pacific, and expand its engineering and technology teams. Further development of its AI security platform is also part of the plan. Engineering capacity matters to this business because its agents are adapted to each customer’s systems, workflows and risks, rather than deployed as an identical service everywhere.

Sales hiring is another priority. In a CTech interview, co-founder and CEO Hagai Shapira said, “Most of the capital will be used to expand our sales team”. That emphasis puts the funding behind both customer acquisition and the staff needed to deliver a service that includes continuing human oversight.

Shapira founded Daylight with Eldad Rudich after both worked at security automation company Torq. Shapira was an early product leader there, while Rudich was its first employee and later led research and development. Their experience is relevant to the operating model now attracting capital: the engineering work continues after agents are built, as customer environments and security needs change.

Agents handle routine investigations; engineers govern the system

The Daylight service model assigns agents triage, enrichment, investigation and response at scale, while security engineers provide oversight and judgment for complex decisions. Engineers also customize the agents to the customer’s environment and keep improving their operation. Human control therefore applies both to the rules under which agents work and to decisions that cannot be reduced to a routine workflow.

An alert may require context from identity, endpoint or cloud systems before anyone can judge its significance. Agents can gather and connect that information, investigate recurring patterns and carry out routine response work. Engineers decide how those capabilities should operate for a particular customer and take responsibility for calls that require expertise about its systems and priorities. The distinction is about accountability during the work, not simply a person reviewing a finished report.

The publicly described model does not specify an approval rule for every possible containment action. The boundary it does establish is that engineers govern the agents and retain judgment calls; the authority assigned to an agent can vary with the customer environment and service arrangement.

Customers choose how to share the security work

Daylight offers fully managed, co-managed and forward-deployed arrangements. Under the fully managed option, it takes on work from alert to resolution. A co-managed customer retains selected detections and processes while Daylight handles other work; a forward-deployed Daylight engineer works within the customer’s team and environment.

Those arrangements change where the customer’s own staff enter an investigation. An organization with an internal security team can keep responsibility for parts of its existing process, while another can hand more of the operation to Daylight. In either case, Daylight’s engineers remain part of the service that builds and governs its agents. The company’s human role is therefore distinct from the customer’s choice about how much work to retain.

Customers can also use Aura, Daylight’s platform, to investigate their environment with natural-language questions, access security data through a centralized data lake and manage operations. In a co-managed arrangement, both teams work in Aura. Shared access makes the division of tasks visible to the customer’s staff, although the service description does not set out every decision an agent may make without approval.

Enterprise AI alerts enter the same workflow

Daylight investigates enterprise AI activity alongside identity and other security signals for customers including Miro.

Daylight’s coverage extends to AI systems used inside enterprises. A SiliconANGLE account describes how alerts involving prompt injection or risky Model Context Protocol integrations can be investigated alongside identity, endpoint and cloud activity; Miro is among the customers using that AI-related coverage. The added signals broaden the work entering the managed detection and response process.

That broader scope also makes the agent-and-engineer division more consequential. Correlating routine signals across systems is work suited to automation, while deciding what a finding means for a particular organization depends on its systems and priorities. Daylight’s engineers are responsible for adapting and governing the agents as the range of activity under investigation expands.

Growth and speed figures remain company claims

The Goldman Sachs Alternatives announcement repeats Daylight’s claims that revenue grew more than 40-fold over the past year and that its service detects and contains threats up to 10 times faster than traditional managed security providers. It gives neither the underlying revenue figure nor a public methodology for the speed comparison. The figures describe the company’s stated performance, rather than independently measured results.

The funding gives Daylight resources to sell and staff its service across the United States. Its expansion will put the central promise of that service under pressure: engineers must continue tailoring and governing agents across more customer environments while retaining responsibility for decisions that require human judgment.