{
  "id": 4845691,
  "title": "Q&A: Serval CEO describes building an ‘AI-native’ alternative to ServiceNow",
  "url": "https://urgent.news/2026/09/01/q-a-serval-ceo-describes-building-an-ai-native-alternative-to",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-01T11:00:00.000Z",
  "source": {
    "name": "Computerworld",
    "slug": "computerworld",
    "url": "https://www.computerworld.com/article/4216413/qa-serval-ceo-describes-building-an-ai-native-alternative-to-servicenow.html"
  },
  "original_language": "en",
  "account": "Since its inception two years ago, Serval has positioned itself as an \"AI-native\" alternative to incumbent IT service management (ITSM) vendors, vowing to streamline the automation of helpdesk workflows. Serval's ITSM platform incorporates a workflow builder that utilizes natural language to construct code-based workflows. Although competitors such as Freshworks and ServiceNow also offer workflow automation tools, Serval asserts that its methodology expedites the automation of recurring tasks, leading to a greater number of tickets being resolved autonomously. Since acquiring clientele such as Spotify, Fox, and Live Nation, Serval has garnered substantial venture capital funding from Sequoia and others, currently boasting an estimated valuation of $1 billion. With its most recent product update, Catalyst, just recently made available to the public, Serval aims to handle a larger portion of the tasks involved in developing and deploying workflow automations. This encompasses analyzing ticket histories and other data to pinpoint processes suitable for automation, subsequently configuring settings and establishing connections to facilitate the implementation of these automations into production. Catalyst additionally enables administrators to devise extended-running background agents capable of identifying issues through historical tickets and other data, suggesting remedies prior to employees even encountering a predicament, as per Serval's claims. Computerworld engaged in conversation with Serval CEO and co-founder Jake Stauch regarding the company's strategy for IT automation and its perception of its capacity to distinguish itself from established ITSM vendors. The subsequent dialogue has been refined and summarized for clarity. What constitute the most significant obstacles IT teams encounter when addressing employee inquiries? \"Conventional ITSM tools are engineered to catalog tickets; Serval is engineered to resolve requests. These tools excel at indicating that work needs to be accomplished, yet they do not alleviate the issue for the employee. The crux of the employee support predicament pertains to automation: someone must devise all the automations required by the employees. \"Traditionally, automations have been constructed within ITSMs as a secondary consideration, employing drag-and-drop workflow builders. These elaborate flow diagrams consume months to develop, at which point someone must uphold them. However, when the business process evolves or an additional approval stage is instituted for password resets, all these disparate workflows necessitate surgical intervention. \"Our differentiation lies in rendering automation more accessible. We contend that the most efficacious approach to crafting automations does not involve a drag-and-drop workflow builder, but rather entrusting AI with the capability to generate code. Our code generation agent assembles workflows for large enterprises to automate not only elementary help desk requests, such as password resets, but also extended employee journeys, including onboarding and offboarding, job transitions, and reporting obligations.\" What are the principal constituents of the Serval platform, and how does Catalyst interrelate? \"Contemplate the system as comprising two agents, albeit it is more intricate than this. \"There exists a help desk agent for employees. An employee may articulate, 'I am barred from accessing my computer,' 'What is the Wi-Fi password?' or 'My laptop is malfunctioning,' through Slack and Teams, over the phone, or via email - and the agent will resolve the matter on the spot. \"We possess a distinct agent, dubbed Catalyst, which constructs these automations. Administrators converse with Catalyst, which can construct any automation, such as password reset workflows, onboarding, offboarding, reporting. It subsequently facilitates the deployment of these automations for either the help desk agent to address employee requests or for other team members to utilize. Catalyst can even engender background agents that conduct investigations, retrieve data from diverse systems, and subsequently furnish reports and propose novel workflows. \"The erstwhile Workflow Builder was more constrained, as it translated your directive into code and fashioned a code-based workflow. You could articulate, 'I necessitate a workflow facilitating the reset of a Microsoft Entra password,' and it could fabricate that. \"With Catalyst, you could articulate, 'Construct a series of password-reset workflows for all the disparate identity providers we employ,' and it would fabricate workflows for Okta, Google, and Microsoft, while accounting for previously established workflows. \"It can also harness other systems to fabricate dashboards. It can scrutinize my ServiceNow data and fashion charts illustrating the information, and subsequently propose automations predicated on its analysis. It can transcend further and fabricate user interfaces. If you assert, 'I require a form for ordering laptops for users,' it can fashion said form, which you can subsequently deploy for equipment acquisitions across the enterprise. What transpires if Serval's AI agents fail to perform as anticipated? What safeguards are in place to impede them from executing unintended actions? \"Serval's architecture is predicated on a partitioning of responsibilities: one agent devises automations, whereas a separate agent employs them. Catalyst is confined to administrators and generates deterministic, code-based workflows. The help desk agent with which end-users interact can invoke solely the workflows previously sanctioned by administrators; it cannot generate new workflows or venture beyond that pre-authorized scope. \"Each workflow carries its own permissions and approval checkpoints that are independent of LLM adjudication. This implies that while the help desk agent engages freely with the employee's request, any action it executes can be traced back to a specific workflow execution, and auditors can peruse both the approvals and the underlying code. \"This amalgamation - of deterministic tools combined with an interstice between the agent that engineers the tools and the agent that utilizes tools - is what enables IT teams to reap the advantages of AI without entrusting extensive system access to a model that renders judgment calls in real time.\" How do ServiceNow and other ITSM vendors incorporate AI agents and agent-building tools into their platforms, and what prevents Serval from emulating their approach? \"ServiceNow and other ITSM vendors are likewise integrating AI agents and agent-building instruments into their platforms. What distinguishes Serval's methodology, and what would impede these enterprises from replicating Serval's strategy? \"While other vendors are augmenting AI agents and agent-building utilities, Serval distinguishes itself through its emphasis on simplifying automation. Serval perceives the optimal avenue for crafting automations as not employing a drag-and-drop workflow builder, but instead permitting AI to pen code. Their code generation agent formulates workflows for substantial enterprises to automate not merely rudimentary help desk requests, such as password resets, but also protracted employee journeys, encompassing onboarding and offboarding, job changes, and reporting requisites. \"In contrast, competitors may employ similar AI-driven workflows, but they may lack Serval's stringent separation of powers, thereby exposing greater vulnerabilities to unintended actions. Serval's dual-agent framework, with the code generation agent confined to admins and the help desk agent limited to executing only pre-sanctioned workflows, furnishes a heightened layer of security. Additionally, Serval's deterministic approach to automations and its comprehensive auditing mechanisms, which trace every action to a specific workflow run, offer a greater degree of accountability compared to their competitors' methodologies.\"",
  "summary": "Since launching two years ago, Serval has positioned itself as an “AI-native” alternative to incumbent IT service management (ITSM) software vendors, promising to simplify the automation of helpdesk workflows. Serval’s ITSM platform includes a workflow builder that uses a natural language interface to build code-based workflows. While vendors such as Freshworks and ServiceNow also have workflow…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}