Daylight Raises $80M From Goldman Sachs to Pair AI Agents With Human Security Engineers
New York-based Daylight, founded in 2025 by two former Torq executives, raised an $80 million Series B led by Goldman Sachs Alternatives, bringing total funding to $120 million. How its hybrid AI-plus-humans model works and what it signals for companies buying security.

NEW YORK — Most cybersecurity startups pitching AI promise to replace people. Daylight is selling a mix: software agents that handle the routine work of watching for attacks, and experienced security engineers who make the calls that require judgment. Investors just put $80 million behind that bet.
The New York-based company announced an $80 million Series B on Nov. 10, led by Growth Equity at Goldman Sachs Alternatives. New investor Okta Ventures, the venture arm of identity company Okta, joined existing backers Bain Capital, Craft Ventures, Vintage Partners and Maple Capital. The round brings Daylight’s total funding to $120 million.
The company didn’t disclose a valuation. Israeli business outlet Calcalist’s Ctech estimated it at roughly $400 million.
What Daylight Does
Daylight works in what the industry calls security operations: detecting, investigating and responding to threats inside a company’s systems. Traditionally that means either building an in-house security operations center, which is expensive and hard to staff, or hiring a managed security provider.
Daylight’s version combines an AI-native platform with its own security engineers. According to the announcement from Goldman Sachs Asset Management, the platform and engineers build, run and customize AI agents for each customer’s systems, workflows and risks. The agents handle day-to-day security operations, while engineers focus on decisions that need a human.
Customers can choose how much they hand over:
- Fully managed, where Daylight runs security operations.
- Co-managed, working alongside a company’s internal team.
- Forward-deployed, with a Daylight engineer embedded in the customer’s organization.
Through its platform, called Aura, customers can investigate their security environment using natural-language questions, keep security data in a centralized data lake and manage operations in one place. Daylight claims it detects threats in seconds and contains them within minutes, “up to 10 times faster than traditional managed security providers.” That is the company’s figure; we haven’t independently tested it.
The Growth Numbers
Daylight says revenue grew more than 40 times over the past year, driven by Fortune 500 companies and AI-native businesses. Named customers include Lovable, LastPass, Rogo, dLocal and Miro. It hasn’t disclosed revenue, but CEO Hagai Shapira told Calcalist the company expects to reach tens of millions of dollars in revenue by the end of 2027.
The company is young. Calcalist reports it was founded in early 2025 and employs 80 people, 43 of them in Israel. Shapira and CTO Eldad Rudich met in an elite Israeli military intelligence unit and later worked at security automation company Torq, where Shapira was an early product leader and Rudich was the first employee and VP of R&D. Angel investors include the founders of Wiz, Cyera, EON and Armis, according to the company.
The money will go toward expanding in the U.S., entering markets across Europe, the Middle East, Africa and Asia-Pacific, hiring security engineers and building out the platform. “Most of the capital will be used to expand our sales team,” Shapira told Calcalist.
Why It Matters Beyond the Fortune 500
Daylight’s customers are mostly large or fast-scaling companies, so this isn’t a product built for the corner shop. But the model says something useful to any growing New York company deciding how to handle security.
First, the hiring problem is the point. Goldman’s announcement describes security engineering talent as expertise “that most companies struggle to recruit and retain.” The pitch is that customers can “run leaner security teams” by letting agents take on routine operations. That’s the same promise many New York startups and midsize firms are now hearing from vendors of every size.
Second, the hybrid approach is a hedge investors seem to like. It’s a theme we’ve seen elsewhere this month: Flatiron-based Hatz AI raised $24 million to help IT providers deliver AI to small businesses, betting that most companies will get AI through people they already trust rather than by deploying it alone.
Questions to Ask Any AI Security Vendor
If you’re evaluating managed security that leans on AI agents, from Daylight or anyone else, a few questions follow directly from how these services are pitched:
- Who makes the final call? Ask which actions agents take on their own and which require a human engineer.
- How fast, measured how? “Seconds” and “minutes” claims should come with definitions and contract terms, not just marketing.
- What do you keep? If the provider centralizes your security data, know where it lives and how you get it back.
- Which model fits? Fully managed, co-managed and embedded models carry different costs and levels of control.
Daylight now has $120 million and an expansion plan. The question for buyers is the one every AI services company is facing: whether the humans in the loop can scale as fast as the software.
