The $11B Market Betting on AI Orchestration Infrastructure

The promise of AI in business has always outpaced the reality. Companies can build impressive AI demos. Getting those demos to actually run reliably inside a real company, at scale, day after day, is a completely different challenge.

That gap is exactly what Orkes is built to solve. The workflow orchestration platform just raised US$60 million in a Series B round led by AVP, with participation from Prosperity7 Ventures (the investment arm of Aramco Ventures), Nexus Venture Partners, Battery Ventures, and Vertex Ventures US. Total funding now sits at approximately US$90 million.

The numbers behind enterprise AI's scaling problem are brutal. McKinsey's State of AI 2025 found that 88% of enterprises now use AI in at least one function, but only 6% see meaningful impact on their bottom line. Two-thirds of companies remain stuck in what analysts call "pilot purgatory," where experiments never graduate to real production systems.

The Infrastructure Problem Nobody Talks About

Most AI coverage focuses on model capabilities. The harder problem is deployment. Moving an AI workflow from a controlled test environment into a live enterprise system means dealing with legacy software connections, inconsistent output quality at volume, security requirements, monitoring gaps, and unclear ownership across teams.

Gartner projects that 40% of enterprise apps will feature task-specific AI agents by 2026, up from less than 5% in 2025. That is a massive buildout happening extremely fast. But 89% of enterprise AI agents never actually reach production, with failed implementations costing companies between US$150,000 and US$800,000 each.

Orkes provides the orchestration layer that sits between AI capabilities and the systems enterprises already run. Think of it as the coordination engine that makes sure AI agents behave predictably, can be monitored, and integrate with existing software without breaking things.

Built on a Netflix Foundation

The credibility behind Orkes matters. The platform is built by the engineers who originally created Conductor, Netflix's internal orchestration tool developed in 2015 to manage the company's explosive growth. Netflix open-sourced Conductor in 2016, and internally its usage has increased five times in recent months. Fortune 100 companies including JP Morgan Chase, Atlassian, Tesla, and American Express run on it today.

Since Orkes raised a US$20 million Series A in early 2024, the company has tripled its enterprise customer base. Clients now include United Wholesale Mortgage, Quest Diagnostics, Twilio, LinkedIn, and Australia's Woodside Energy.

The results at United Wholesale Mortgage are telling. Andy French, AVP of Platform Automation at UWM, said after deploying Orkes: "Our DevOps Architects don't have to spend 95% of their time managing Conductor. They can focus on creating new services and features." That shift, from infrastructure firefighting to actual product work, is the practical payoff of reliable AI deployment.

What This Means for Business Leaders

The AI orchestration market is projected to grow from US$11 billion in 2025 to US$30 billion by 2030. That growth is not driven by excitement over AI models. It is driven by the hard operational reality that getting AI to work reliably inside a business requires dedicated infrastructure.

For executives outside the technology sector, the Orkes funding round is a useful signal. The enterprise AI conversation is shifting. The question is no longer whether your organization should use AI. It is whether you have the infrastructure to deploy it in a way that actually changes business outcomes. Only 5% of companies currently achieve substantial AI returns at scale. The ones that do average 1.7x ROI compared to peers still running pilots.

As Jeu George, co-founder and CEO of Orkes, put it: "Developers need orchestration, controls, and visibility to run advanced AI and agentic systems with confidence."

The funding round signals investor conviction that this infrastructure layer is where the real enterprise AI value will be built.