Serval CEO: AI-Native ITSM Built to Resolve, Not Just Track Tickets

Serval positions code generation as the path to faster IT automation

Serval, a two-year-old IT service management platform valued at around $1 billion, is challenging incumbents like ServiceNow with a fundamentally different approach to workflow automation. Rather than drag-and-drop builders, the company uses AI agents that write code from natural language prompts.

CEO and co-founder Jake Stauch argues that traditional ITSM tools excel at tracking tickets but fall short at actually resolving employee requests. "The problem with employee support is really around automation: somebody has to go and build all the automations employees need," Stauch told Computerworld in a recent interview.

Serval's platform centers on two distinct agents. One handles employee-facing requests through Slack, Teams, email, or phone. The other, called Catalyst, builds the automations themselves. Admins describe what they need in plain language, and Catalyst generates code-based workflows—whether for password resets, onboarding sequences, or license reclamation.

The company's latest Catalyst update, released last week, goes further by analyzing ticket histories to identify automation opportunities, then configuring settings and integrations to deploy those workflows. It can also create persistent background agents that detect issues like security vulnerabilities or failed workflows before employees encounter them.

Why it matters

Serval's architecture illustrates a strategic bet: that enterprises will favor platforms designed for AI from the ground up over legacy systems retrofitted with agent capabilities. The company's customer list—including Spotify, Fox, and Live Nation—suggests demand exists for ITSM tools that prioritize automation speed over configurability. For IT leaders evaluating service management platforms, the question becomes whether existing vendor roadmaps can match the velocity of purpose-built alternatives, or whether technical debt in decades-old architectures creates an insurmountable gap.

Separation of powers as a safety model

Stauch emphasized that Serval's security model relies on architectural separation. Catalyst, restricted to admins, writes deterministic workflows. The employee-facing agent can only execute pre-approved automations—it cannot create new ones or access systems outside its defined scope. Every workflow carries its own permissions and approval gates that aren't subject to large language model judgment calls.

The incumbent challenge

When asked how Serval differentiates from ServiceNow's own AI agent tools, Stauch pointed to architectural constraints. "It's that their existing customers depend on years of customization built on the old architecture, and incumbents can't tear that out without breaking what those customers already rely on," he said.

Serval can initially deploy on top of existing ITSM platforms, syncing data back to ServiceNow or similar tools. But Stauch noted that no customer wants that arrangement long-term. Many migrate fully when contract renewals approach, replacing rather than supplementing their incumbent platform.

Beyond IT departments

Despite marketing to IT buyers, Serval's typical customer deploys the platform across 13 departments, including HR, finance, legal, and security. Stauch said IT accounts for a minority of teams actually using the system—the department serves as the entry point, but automation needs extend throughout the organization.

On workforce impact, Stauch reported that customers redeploy rather than reduce IT staff. Freed from repetitive ticket work, admins build automations for other business functions. One customer, Perplexity, reclaimed one to two hours per day per admin through automation, time that gets reinvested in building tools for other departments.

These details were first reported by Computerworld in an interview with Stauch.