NFON launches AI Agent Studio for building agents from plain-language descriptions

NFON AG, the Munich-based business communications provider, launched AI Agent Studio on November 12, adding a conversational way to build AI agents to the visual approach it already offers. The company says the product is available now, and it is aimed squarely at European small and mid-sized businesses and the partners that serve them.

The design premise is that the hard part should not be modelling. Users describe in natural language what an agent should do, how it should behave and which constraints it must respect, and the system generates an initial working version that can be tested immediately and refined through further dialogue.

Two development paths sit inside one workspace. Agent Mode is the natural-language route: the first solution is generated automatically, users can add requirements and request changes, and they review proposed edits before accepting them. Built-in controls keep changes traceable and reversible, which NFON frames as the oversight that more demanding enterprise deployments require.

Flow-Based Mode returns to visual configuration. Users lay out workflows and set each step by hand, specifying exactly how a flow is controlled and what the agent must do under defined conditions. NFON describes this as precise, rule-based control that complements the conversational path; both modes can draw on knowledge sources, tools and extended workflows.

The reference use case is customer service. An agent answers FAQs from stored material on products, services or internal processes, receives an enquiry, asks for missing information and routes the details to the right team. Capabilities extend through tools and interfaces such as CRM or appointment systems, with custom integrations available and NFON partners or Professional Services handling heavier projects.

Chief executive Andreas Wesselmann casts the studio as the foundation for growing with customer requirements and covering more operational tasks, strengthening NFON's role as a long-term AI partner for European SMEs while providing the base for more complex work in larger organisations. Jana Richter, executive vice president for engineering, AI and innovation, put the goal plainly: to democratise advanced AI automation, so that ideas rather than technical barriers determine what a company can achieve.

For NFON's channel, the commercial logic is speed. The company notes that its voicebots already take insurance claims, process meter readings for energy suppliers and pre-qualify enquiries in tradespeople's workshops, and that local government agents can answer questions on public services, departmental responsibilities and opening hours. Partners can turn a specific customer requirement into a demonstrable prototype and iterate with the client instead of opening a long development project.

NFON is backing the launch with training and certification programmes, implementation examples and customer-consulting material, with distribution partners owning outreach, advice and marketing and, depending on the project, technical implementation. The open question is adoption: whether European SMEs will hand process automation to agents described in plain language by the people who do the work.