News
CBTS Takes Its Agentic AI Playbook to the Mid-Market
Large enterprises have dominated the early push to deploy agentic AI across the workplace. Technology services company CBTS thinks the next opportunity lies further down the market.
CBTS on Wednesday unveiled an agentic AI deployment blueprint aimed at mid-market companies, which it defines as organizations with between $300 million and $3 billion in annual revenue. The company developed the approach rolling out Anthropic's Claude Enterprise across its 2,300-employee workforce. The idea is to give companies without the resources of the largest enterprises a path from experimenting with AI to deploying it across business operations.
"With CBTS, all companies can now access world-class AI knowledge, security, consulting frameworks, and analytics once reserved for large corporations," CBTS CEO Kristin Russell said. "The goal is to give mid-market companies across North America the tools to compete on equal footing with companies ten times their size."
From AI Pilots to Agents
CBTS calls its goal "100% agentic delivery," meaning agentic AI is incorporated across business functions including sales, operations, security, IT, and service delivery. That's an ambitious target, and the company did not provide financial or productivity results showing what its own deployment has delivered. But the approach illustrates how the enterprise conversation around AI is shifting.
Rather than simply giving employees access to a chatbot, organizations are increasingly looking at how AI can become part of existing workflows and interact with other business systems. That introduces additional questions around infrastructure, security, governance, and human oversight.
CBTS says it encountered those issues during its own Claude deployment. Employees now use Claude.ai for research and drafting, Claude Code for engineering and technical work, Cowork for cross-functional projects and Claude Design. CBTS said security and governance were a board-level mandate as it expanded the technology across the company.
The lessons from that rollout now underpin CBTS Forge AI, the company's practice for helping customers deploy AI.
One Model won't Necessarily do Everything
Despite building its internal deployment around Anthropic, CBTS isn't proposing an Anthropic-only strategy for customers.
Its Forge AI approach is multi-model by design, with organizations able to select different models based on factors including performance, cost, security, and data-sovereignty requirements.
That's significant as companies consider deploying agents for increasingly specialized jobs. The model best suited to one workflow may not offer the right combination of cost, performance, or security for another. CBTS's blueprint therefore stretches beyond model selection.
The company says it will provide advisory services to identify potential AI use cases, technical expertise to build and govern agentic systems, and infrastructure and managed services to operate them in production. Human workers remain part of that equation. CBTS said its approach prioritizes employee reskilling and human oversight as agent use expands.
Bringing Agentic AI to the Mid-Market
The larger bet is that mid-market organizations want many of the same AI capabilities being pursued by major enterprises but don't necessarily have the internal engineering, security, and governance resources required to build them independently.
CBTS argues that those constraints have made large-scale AI adoption disproportionately accessible to companies with enough money and technical resources to experiment, make mistakes, and rebuild.
Its answer is essentially to package the experience gained from its own deployment into a repeatable model for smaller organizations. Whether businesses ultimately embrace CBTS's vision of "100% agentic delivery" remains to be seen. The company hasn't provided evidence that every business function needs an AI agent, and its announcement doesn't quantify the business gains from its own company-wide Claude deployment.
But the launch also indicates that the agentic AI push is broadening beyond the largest companies. As it does, the challenge for organizations with fewer resources may increasingly become not whether they can access powerful AI models, but whether they have the infrastructure and controls needed to put them into production safely.
"Every mid-market CISO is asking the same question: How do I use AI to do more with less, without opening a door I can't close?" CBTS CISO Chris DeBrunner said.