任何在大型企业工作过的人都知道这套流程。你为复杂的软件问题提交一张IT工单，然后只能苦等。
Anyone who has ever worked at a large enterprise knows the drill. You submit an IT ticket for a complex software issue, and then you wait.
第一代AI聊天机器人已经成功实现了一级支持中最简单工作的自动化，例如密码重置和基础软件审批。但过去，只要请求需要真正的问题解决能力，它们就会立刻碰壁。
First-generation AI chatbots have successfully managed to automate the low-hanging fruit of Tier-1 support, like password resets and basic software approvals. But they historically hit a brick wall the second a request requires actual problem-solving.
如今，AI原生服务台初创公司Risotto正试图解决这一难题。该公司今天宣布推出其“自主IT工程师”。这是一款专为处理复杂且模糊的二级支持请求而打造的新一代AI副驾驶。
Now an AI-native helpdesk startup, Risotto, is trying to solve that headache. Today the company announced the launch of its ‘Autonomous IT Engineer’. It is a new AI co-pilot designed specifically to tackle complicated, ambiguous Tier-2 requests.
传统IT自动化的根本缺陷在于，无法随着复杂程度的提升实现规模化。当未曾见过的边缘情况出现时，仍需人类工程师介入。他们调查日志、诊断问题，并手动实施修复。
The fundamental flaw of traditional IT automation is that it doesn’t scale with complexity. When an unseen edge case pops up, human engineers have to step in. They investigate logs, diagnose the problem, and manually deploy a fix.
真正的痛点不只是劳动成本，而是组织知识的流失。那些来之不易的解决方案通常会消失在一张张已关闭的工单记录中。这意味着下次问题再次出现时，团队仍需付出完全相同的人工精力。
The real pain point isn’t just the labor, it’s the lost institutional knowledge. Those hard-won resolutions usually disappear into closed ticket histories. That means teams will need the exact same manual effort the next time the issue arises.
Risotto正在正面应对这一生命周期缺陷，打造能够自行编写操作手册的智能体。该公司表示，其全新的“自主IT工程师”并非只是一个对话机器人，而是一名数字员工，能够调查此前从未遇到的问题。通过分析实时用户上下文和系统日志，该智能体可以诊断问题，并提出分步骤修复方案，随后在人工审批参与下执行。
An agent that writes its own runbooks Risotto is attacking this lifecycle flaw head-on. The company says its new Autonomous IT Engineer doesn’t just act as a conversational bot. It operates as a digital employee capable of investigating entirely new problems. By analyzing real-time user context and system logs, the agent diagnoses the issue and proposes a step-by-step fix. It then executes it via human-in-the-loop approval.
关键在于，Risetto在设计系统时赋予了它学习能力。一旦解决了一个新问题，Risotto就会自动将这一成功的修复方案转化为可重复使用且能够自我改进的操作手册。
Crucially, Risotto designed the system to learn. Once it solves a novel problem, Risotto automatically converts that successful fix into a reusable, self-improving runbook.
“每个AI帮助台都能处理密码重置请求，但IT领域真正的瓶颈在于梳理并自动化此前没有文档记录的复杂运营工作流，”Risotto首席执行官兼联合创始人阿伦·索尔伯格说。“我们的自主IT工程师不仅会解决这些复杂问题，还会将每个解决方案转化为可重复使用的自动化流程。这些智能体不只是调用工具，还会构建自己所需的工具。”
“Every AI helpdesk can deflect password resets, but the real bottleneck in IT lies mapping out and automating complex operational workflows that haven’t been previously documented,” said Aron Solberg, CEO and Co-Founder of Risotto. “Our Autonomous IT Engineer doesn’t just resolve these complex issues; they convert each resolution into a reusable automation. The agents don’t just use tools, they build the tools they need.”
挖掘历史工单 为确保平台无需从零开始，Risotto表示其内置“历史工单挖掘”功能。该功能会分析公司过去90天的工单历史，自动发现并生成新的自动化工作流。用户也可以用自然语言描述所需的自动化流程，随后Risetto会构建并测试这些流程，并通过API或MCP端点连接现有企业工具。
Mining past tickets To ensure the platform doesn’t start from zero, Risotto says it comes equipped with a ‘Past Ticket Mining’ feature. It analyzes 90 days of a company’s ticket history to automatically surface and generate new automation workflows. Users can also describe the automations they want in natural language. Risotto then builds and tests them, connecting to existing corporate tools via API or MCP endpoints.
该平台还在突破被动支持的局限。Risotto可以在几分钟内生成基于实时数据的定制仪表板和报告，从而无需另行向BI或工程团队提交工单。IT团队还可以为定时或事件驱动的工作流设置主动触发器，例如系统健康检查或员工入职流程。这使IT部门的角色从被动救火转向主动运维。
The platform is also pushing beyond reactive support. Risotto can generate custom, live-data dashboards and reports in minutes. That bypasses the need to file separate tickets with a BI or engineering team. It also allows IT teams to set up proactive triggers for scheduled or event-driven workflows. Examples include system health checks or employee onboarding sequences. That shifts the IT department’s role from reactive firefighting to proactive operations.
客户与成果 这种智能体驱动的IT服务管理（ITSM）方式已逐渐获得市场认可，Risotto称目前的客户包括Gusto、Instacart、Docker、Retool、ThoughtSpot、Patreon和Ironclad。
Customers and results This agentic approach to IT service management (ITSM) is already gaining traction, and Risotto claims current customers include Gusto, Instacart, Docker, Retool, ThoughtSpot, Patreon, and Ironclad.
这些早期采用者报告的效率提升幅度十分显著。Risotto表示，客户的平均首次响应时间已降至4秒以内，自动解决率介于40%至90%之间。更令人印象深刻的是，在与Ironclad合作的一项案例研究中，该公司声称实现了91%的自动解决率。Vidyard称，其一级请求在30天内的自动化率达到80%。在Gusto，该平台首日解决率达到53%，据估计节省了11.4万小时的支持等待时间。
The efficiency gains reported by these early adopters are steep. Risotto says customers are seeing average first-response times drop to under four seconds, with auto-solve rates between 40-90%. Most impressively, in a case study with Ironclad, the company claims to have achieved a 91% autosolve rate. Vidyard claimed an 80% automation rate of tier-1 requests within 30 days. At Gusto, the platform achieved a 53% resolution rate on day one, with an estimated 114,000 hours of support wait time saved.
Jobber是Risotto的另一家客户。该公司IT运营经理迈克·史密斯表示：“Risotto的人工智能支持系统几乎无需任何设置，开箱即用便产生了巨大影响。员工如今可以迅速高效地获得答案，其精准程度与我们的专业人员别无二致。”今年早些时候，该公司宣布完成由Bonfire Ventures领投的1000万美元种子轮融资，645 Ventures、Y Combinator、Ritual Capital、Surgepoint Capital以及Dropbox和HelloSign的前高管也参与了投资。该公司目前似乎仍拥有充足的资金。
“Risotto’s AI-powered support system had a massive impact straight out of the box and with very little setup,” noted Mike Smith, IT Operations Manager at Jobber, another Risotto customer. “Employees now get fast, efficient answers with the same precision as one of our professionals.”Earlier this year, the company announced a $10M seed round led by Bonfire Ventures, along with 645 Ventures, Y Combinator, Ritual Capital, Surgepoint Capital, and former executives from Dropbox and HelloSign. The company appears to still be well capitalized.
Risotto押注的是，随着企业技术栈日益碎片化和复杂，企业将乐于投入资金，通过自动化消除内部支持中的摩擦。Risotto认为，将一次性的手动解决方案转化为永久性的自动化工作流后，IT的未来不仅在于更快地回答问题，更在于确保同一个问题永远不必回答第二次。
Risotto’s bet is that as the corporate tech stacks grow increasingly fragmented and complex, companies will happily pay to automate away the friction of internal support. By turning one-off human fixes into permanent, automated workflows, Risotto is saying that the future of IT isn’t just about answering questions faster, it’s about ensuring the same question never has to be answered twice.