Wonderful raises $550M for an AI strategy built around hiring humans

Why it matters

Wonderful's $5B valuation rests on a founder-led bet that enterprise AI will be won through local engineering and integration, even if that produces a far heavier cost base than conventional SaaS.

Bar Winkler and Roey Lalazar raised $550 million for Wonderful on September 2, giving their enterprise AI operation a $5 billion valuation and fresh capital to expand a deployment model built around putting engineers inside customer organizations.

TechCrunch reported that Insight Partners led the Series C. Salesforce was the only new investor, joining returning backers Index Ventures, IVP, Vine Ventures, 9Yards and Bessemer Venture Partners.

Wonderful says it has reached 650 employees and more than 35 markets since it was founded in 2025. The latest financing follows a $34 million seed round in July 2025, a $100 million Series A that November and a $150 million Series B in March 2026. Those rounds add up to about $834 million in announced funding in roughly 14 months.

The $5 billion valuation also marks a steep reset from the $2 billion price attached to Wonderful's Series B less than six months ago. The new round gives Winkler and Lalazar the money to test whether a people-intensive implementation strategy can scale as quickly as Wonderful's investors expect.

Two founders built around the last mile

Winkler arrived at Wonderful with direct experience selling workflow software into finance departments. He previously co-founded Approve.com, which handled procurement requests, approvals, purchase orders and vendor onboarding. Tipalti acquired Approve.com in April 2021 for an undisclosed price, and Winkler subsequently ran procurement at Tipalti through 2024.

Before Approve.com, Winkler worked at ironSource and helped launch its mobile product. Lalazar previously co-founded Kaps, an AI localization operation, and worked in software engineering and cybersecurity.

Their backgrounds explain Wonderful's original wedge. Wonderful started with customer-facing agents across voice, chat and email, concentrating on markets where language, dialect, regulation and local business practices made generic English-first systems difficult to deploy. Index Ventures said when it led Wonderful's seed round that the initial product was already handling interactions for 15 large customers.

Wonderful has since widened the pitch. Its current product, marketed as the Wonderful AI OS, connects agents, workflows, applications, enterprise data and existing integrations. Wonderful says customers can use different foundation models for different tasks and deploy the software in their preferred cloud or on their own infrastructure.

That model-agnostic architecture gives customers room to switch model providers as performance and prices change. It also positions Wonderful above individual customer-service agents, where vendors including Sierra, Decagon, Intercom and Salesforce already compete for enterprise budgets.

The product comes with engineers

Wonderful's defining product decision is its use of forward-deployed engineers. These employees work alongside customers, mapping workflows, connecting legacy systems and getting an initial use case into production. Wonderful says the engineers then transfer enough knowledge for customers to build and operate additional deployments internally.

Winkler described the economics of that strategy plainly in a June interview with Globes: "The product itself isn't worth much if you don't connect it to dozens of systems within the enterprise."

That approach addresses a real constraint in enterprise AI. A capable model does little on its own when the useful data sits across old databases, call-center software, internal approval systems and industry-specific applications. Wonderful is selling the engineering work required to connect those systems alongside the software that runs on top of them.

The trade-off appears in the headcount. Wonderful now reports 650 employees worldwide. The company says it plans to expand toward a 1,000-person global workforce.

A separate hiring target requires careful distinction. Lalazar told ETtech in August that Wonderful could hire as many as 1,000 forward-deployed engineers in India over 10 to 12 months. That India-specific figure is separate from the company's previously stated worldwide headcount plan.

The Series C will fund product development and additional deployment teams, according to Wonderful's September 2 announcement. Insight is underwriting that hybrid model for a third time. Index led the seed and Series A rounds, while Insight participated in the Series A before leading both the Series B and Series C.

The valuation runs ahead of the published numbers

Wonderful's revenue disclosures remain management claims, and their rapid progression helps explain the size of the wager. Globes reported that annual recurring revenue rose from about $1 million in August 2025 to $7 million at the end of that year and $18 million in the first quarter of 2026.

In August, Lalazar told ETtech that annualized revenue was approaching $100 million. Even against that newer figure, a $5 billion valuation would equal roughly 50 times annualized revenue. Wonderful has not published enough detail on retention, customer concentration, pricing or gross margins to assess how much of that revenue carries software-like economics after the cost of local engineering teams.

Wonderful's expansion provides some operating evidence beyond the financing cadence. Wonderful says its platform now operates in more than 35 markets, up from more than 30 at the Series B.

The capital allows Winkler and Lalazar to keep hiring ahead of the revenue curve. It also commits Wonderful to proving that forward deployment can become a repeatable distribution system rather than a growing implementation burden.

Enterprise customers are asking vendors to connect models to real workflows, maintain those systems and satisfy local compliance requirements. Wonderful has raised enough money to put engineers wherever that work happens. The $5 billion question is how much of their knowledge Wonderful can eventually turn into software.