Serval AI agents just earned one of the loudest votes of confidence yet in the race to automate enterprise work. The San Francisco company has raised 150 million dollars in a round led by Sequoia, tripling its valuation to 3 billion dollars, according to reporting by Reuters. The deal lifts Serval's total funding to 277 million dollars and sharpens a direct challenge to established workplace software providers.

Serval AI agents are built that handle workplace processes, starting with IT operations such as managing employee access, troubleshooting problems, and tracking software and hardware assets. According to Reuters, the new funding will support an expansion of Catalyst, the company's product for building and running automated workflows. The pitch is blunt: Serval AI agents are not designed to sit on top of legacy platforms. It wants to replace them.

That ambition puts Serval squarely in the middle of the year's biggest enterprise software argument. Incumbents such as ServiceNow and Workday have spent years embedding automation into their platforms, while a wave of agent startups argues that a fresh architecture can do the job faster and cheaper. Serval's valuation jump, from 1 billion dollars in December to 3 billion dollars now, suggests investors are increasingly willing to bet on the challengers.

Serval AI agents take aim at legacy platforms

According to Reuters, Serval co-founder and chief executive Jake Stauch described the company's product as a true alternative replacement for platforms like ServiceNow, rather than an additional layer of AI spending on top of an existing system of record. That framing matters because it defines the sales fight: Serval is not asking IT teams to buy one more tool, but to move core workflows off the systems they already pay for.

Stauch pointed to electric vehicle maker Lucid Motors as evidence, telling Reuters that Lucid migrated from ServiceNow to Serval in 45 days. A migration that fast, if it holds up across larger and more complex deployments, would undercut one of the strongest defences incumbents have: the switching cost of ripping out entrenched software. Enterprises have historically tolerated clunky workflows precisely because migration looked painful, slow, and risky.

The timing also fits a broader pattern across the agent economy. Recent coverage of agentic platforms shows funding flowing to startups that automate specific operational jobs end to end, from finance operations to frontline hiring. The Serval AI agents bet is that IT operations, with its repetitive access requests, tickets, and asset tracking, is one of the ripest places for agents to prove they can carry real work rather than merely suggest it.

Guardrails and human approval remain central

The most interesting part of the Serval AI agents story may be how it handles trust. According to Reuters, Stauch said Serval separates the agent used by administrators to build tools from the agent employees use to access approved workflows. Permissions and approvals stay under the control of IT and security teams, and customers still generally want a human to approve sensitive actions such as password resets, application access, and changes to sensitive systems.

That design reflects a lesson the whole agent industry is learning at speed: autonomy without guardrails does not sell into the enterprise. Companies will hand agents routine work, but they want clear boundaries, audit trails, and a human in the loop for anything irreversible. Serval's architecture, with distinct builder and employee-facing agents, is an attempt to give IT leaders both automation and control in the same package.

Previous investors including Elad Gil, First Round, and General Catalyst joined the latest round, according to Reuters, a sign that early backers are doubling down rather than taking profits at the higher valuation. With 277 million dollars raised in total, Serval AI agents now have the capital to expand Catalyst, hire aggressively, and compete for the large enterprise deals that will decide whether agent-native platforms genuinely displace incumbents or settle into a supporting role alongside them.

The stakes go beyond one company. If Serval AI agents and their peers can migrate customers in weeks rather than years, the enterprise software market could shift faster than incumbents expect. If migrations stall on security reviews, integrations, and change management, the valuation will look like another peak in a frothy cycle. Either way, the next few quarters of Serval deployments will be a useful test of how far AI agents have really come, and how much of the workplace they are ready to run.

For more on the agent economy, see ServiceNow Google Cloud Deal Brings Agentic AI to Workflows and Casap Agentic AI Raises $56M to Automate Finance Work, and read the original reporting at Reuters.