Overview
As an AI and data-led consulting firm, evolv needed to walk the talk — adopting AI natively across its own business before asking clients to do the same. The challenge was two-fold: driving genuine adoption across a distributed workforce, and embedding AI into billable client delivery in a way that produced measurable, defensible outcomes.
evolv partnered with Anthropic to deploy Claude Enterprise company-wide as the foundation of its internal AI transformation initiative (branded “nAtIve”), while simultaneously using Claude Code to accelerate complex client engagements. The firm went further by engineering two production-grade agentic applications on the Anthropic SDK, proving expertise as a builder on Claude — not only as a consumer of it.
action plan
evolv embedded Claude across its entire business — internal operations, client delivery, and custom application development — to prove that an AI-first consulting model could produce measurable outcomes both inside the firm and on client engagements. The primary goal was to move beyond AI adoption as a talking point and into AI as a core delivery capability, culminating in two production-grade agentic applications built directly on the Anthropic SDK.
Steps
1
Deployed Claude Enterprise company-wide across internal operations, go-to-market, and partnerships functions as the foundation of evolv’s AI transformation initiative.
2
Integrated Claude Code into client-facing engineering workflows across healthcare and financial services engagements, driving significant reductions in development time.
3
Engineered and shipped two production-grade agentic applications on the Anthropic SDK — the Rapid Prototyping Engine (RPE) and Project Ultra — forming an end-to-end pipeline from client discovery to working demo.
Results
8,000+ hours saved
during Anthropic’s private preview period, with Claude now formalized across evolv’s entire delivery model.
96–99% reduction in development time
on a healthcare client database migration, with similar gains on financial services engagements.
Two production agentic applications shipped
RPE generated 59 client demos in its first quarter at a fraction of the previous manual effort; Project Ultra compressed four-week discovery engagements to two days.
Conclusion
“We didn’t want to just recommend AI to our clients. We needed to prove it. Claude became the backbone of how we work internally and how we deliver for clients. The results speak for themselves.” – Trent Foley, evolv CTO