A new joint venture called Ode with Anthropic is making a bold wager: that a small team of skilled engineers can accomplish what has traditionally required large armies of consultants. The venture is focused on embedding forward-deployed engineers directly inside enterprise companies to help them build and scale artificial intelligence solutions.
The initiative has attracted backing from a notable group of investors and partners, including Anthropic, Blackstone, Hellman & Friedman, and Goldman Sachs, among others. By placing technical talent directly within client organizations, Ode aims to bridge the gap between experimentation and real-world deployment.
The Origins: Fractional AI Becomes the Core
Ode's foundation was built through acquisition. Earlier this year, the venture acquired Fractional AI, an applied AI services startup co-founded by Chris Taylor and Eddie Siegel. That startup now serves as the operational core of the new entity, and both founders have taken on leadership roles at Ode.
The acquisition strategy underscores a broader belief driving the venture: that applied AI services — not just standalone products — represent a critical and rapidly growing category in the technology landscape. Rather than selling software licenses and leaving clients to figure out implementation, Ode's model centers on hands-on engineering work conducted inside the enterprise.
Why Enterprise AI Pilots Stall
Speaking on TechCrunch's Equity podcast, Taylor and Siegel joined host Rebecca Bellan to discuss a persistent problem in the enterprise AI space: the large share of pilot projects that never reach production. Many companies launch AI initiatives with enthusiasm, only to struggle when it comes to integrating models into live business processes, ensuring reliability, and demonstrating measurable returns.
Ode's leadership believes that the missing ingredient is often deep, embedded technical expertise. Forward-deployed engineers who work closely with internal teams can navigate the complexities of data pipelines, model customization, and operational constraints that typically derail pilots. This hands-on approach, they argue, is what separates experiments from deployed systems.
