Transparent pricing.
No per-seat fees.

AIgent is a sovereign AI agent platform. You deploy pre-configured agents, connect them to your data, and use them anywhere: web widget, Slack, Telegram, or your own API. Unlike a plain chatbot, an agent does real work: it plans, calls tools and skills, works in its own sandboxed workspace, and remembers context across sessions. Every step, from the plan through the tool call to the result, stays visible and auditable. The platform suits any company, from SMEs to regulated banks, because you decide hosting and model per agent. The key difference: AIgent combines the autonomy of a true agent platform with Swiss data residency and complete traceability, instead of forcing you to choose between power and control.

AIgent is a GDPR- and Swiss revFADP-compliant AI agent platform hosted in Switzerland. By default all data stays in Swiss data centers, independent of US hyperscalers and outside the US Cloud Act. You can also run AIgent fully on-premise, up to air-gapped, so no data leaves your infrastructure. Data export and account deletion are built in, and every operation is recorded in the audit trail with model, tokens, cost, and time. Metrics are produced over agents and tasks, not over individual people, which rules out employee surveillance from the start. The platform is designed for revFADP, GDPR, the EU AI Act, and DORA. The concrete difference from many vendors: data residency and model choice are not a paid add-on but a core principle that you control per agent.

With AIgent you build an AI agent in under 15 minutes, no coding required. You define the task and guardrails, choose a model and security level per agent, connect tools like Google or Microsoft 365, and add tested skills from the catalog. The agent runs in an isolated Swiss workspace and is instantly available via web widget, Slack, Telegram, or API. Instead of starting from scratch, you begin from a template and adapt it to your use-case. IT and the business configure centrally, while the department uses the finished agent with no further setup. If you need custom logic, you bring skills in the open Claude Code or Codex format. The difference from generic builders: Swiss hosting, model choice, and complete traceability are there from the first agent, not only in a later enterprise tier.

Yes. AIgent is a no-code platform for AI agents. Business teams pick a pre-configured agent and start without writing a single line of code. You set the task, system prompt, and guardrails through the interface, and connect tools with a click. IT and compliance centrally configure model, tools, and permissions and roll the agent out under control. If you need custom logic, you bring tested skills in the open Claude Code or Codex format, or write your own in a folder with a skill.md. This keeps the platform usable for non-technical staff and extensible for developer teams. The practical difference: you need no engineering team of your own to become productive, yet the full path to custom code stays open whenever a use-case calls for it.

AIgent is sovereign and a real agent platform at once. US providers like ChatGPT Enterprise or Microsoft Copilot remain subject to the US Cloud Act, even in an Azure Swiss region, and bill per seat. AIgent runs in Swiss data centers or on-premise, with model choice per agent, tested skills, and a complete audit trail, usage-based with no per-seat fees. Where many tools only chat, an AIgent agent does real work over tested code and delivers reproducible results instead of estimated text. You are not tied to a single model provider but choose, by data sensitivity, a model in-house, Swiss-hosted, or international. The decisive difference is the combination: sovereignty, true agent capability, and full traceability do not come together at the named vendors, but they do at AIgent.

Any US entity can be compelled to hand over data even on Swiss or EU servers. The US Cloud Act and comparable laws apply at the company's domicile, not at the location of the data center. Even an Azure region in Switzerland is subject to it, because Microsoft is a US company. A contractual assurance does not change this as long as the provider is under US jurisdiction. AIgent runs off US hyperscalers, in our Swiss cloud or entirely on your premises, up to air-gapped. That puts data sovereignty with you, not with a foreign parent company. The concrete difference: sovereignty comes from the operator structure and the hosting, not from a label on a region of a US provider.

AI rarely fails on technology, mostly on trust. Trust comes from sight. Every agent shows step by step what it does: plan, model, tool call, result. Instead of an opaque black box that only returns an answer, you see the whole path to it and can review it before you trust it. Every interaction is captured with model, tokens, cost, and time and is auditable, complete and without reference to individuals. Through one dashboard you follow all agents together: what they deliver, which model they use, how long they take, and what they cost. It is exactly this visibility that IT and compliance need to sign off on an agent. The difference from many vendors' black box: with AIgent, traceability is built in, not reconstructed after the fact.

Skills are tested code, not just text. Same input, same output. Calculations, reports, and documents are produced reproducibly and verifiably, instead of being estimated by a language model. Where a plain chatbot would plausibly guess a number, a skill runs the code behind it and returns a deterministic, traceable result. Each skill runs in its own isolated workspace, separated from other agents and data. You install skills from the catalog, bring existing Claude Code or Codex skills, or write your own in a folder with a skill.md. Because the format is open, there is no proprietary lock-in. The decisive difference: for anything that has to be exactly right, such as tariffs, deadlines, or reports, you rely on tested code instead of a guess.

The model follows the sensitivity of the data: an on-premise model for the most sensitive data, Swiss-hosted models for confidential business data, or powerful international models like GPT, Claude, and Gemini for non-critical tasks. You choose per agent instead of committing to a single provider. The most sensitive data never leaves the building, because the corresponding model runs on your own infrastructure. Open-source models can be connected as well when a use-case requires it. Because AIgent is model-independent, you benefit from new models as soon as they become available, without switching platforms. The practical difference: you steer the level of data protection and performance deliberately per agent, instead of routing every operation through the same model.

Yes. On-premise customers run the platform on their own Kubernetes cluster using our Helm chart, up to air-gapped. It is the same product as in the Swiss cloud, only entirely in your own environment. In this mode neither data nor models leave your infrastructure, and you integrate Entra ID, single sign-on, and role synchronization from your existing directory services. On-premise models process the most sensitive data locally, while for non-critical tasks you can still choose a different model per agent. On-premise is licensed per installation, not per seat. Contact sales for options, licensing, and the technical alignment. The difference from the cloud: full data sovereignty and operation in your hands, with identical functionality and the same traceability.

In Swiss data centers by default, under revFADP and GDPR. There is no transfer to the US unless you deliberately connect an international model for a specific agent. Even then you control per agent which data reaches which model at all, so the most sensitive content stays in Switzerland or in-house. If you run on-premise, your data lives solely where you decide, up to an air-gapped environment with no internet connection. Data export and account deletion are built in, so you can take your data with you or remove it at any time. The concrete difference from US clouds: the location is not just a region in a hyperscaler's console but tied to an operator under Swiss law, outside the US Cloud Act.

Yes. Swiss data residency, on-premise, model choice per agent, a complete audit trail, and data export are built in, designed for revFADP, EU AI Act, and DORA. Banks, insurers, and the public sector deploy the platform exactly where supervisors and auditors require traceable processes. The most sensitive data is processed by an in-house model, and every operation is captured with model, tokens, cost, and time and is verifiable. Metrics are produced over agents and tasks, not over people, which respects the requirements of employee data protection. Bring your compliance team in early; sales joins the technical discussion and the alignment with your supervisor. The difference: compliance here is not a later add-on but part of the architecture.

Every AI interaction is logged with model, tokens consumed, cost, and response time. On top of that comes the agent's own flow: which plan it formed, which tool it called, and which result it produced. This makes every operation fully reconstructable and auditable after the fact. Metrics are produced over agents and tasks, not over individual employees, so no personal surveillance is possible. Conversation history is stored per session, and data export and account deletion are built in. You therefore keep control over what stays stored and what gets deleted. The difference from many black-box services: the audit trail is complete and part of the platform from the start, instead of being pieced together from scattered logs.

Yes. Everything you can do in the UI is also available through our API: create agents, send messages, trigger skills, upload knowledge, pull usage. This lets you embed agents directly into your existing systems and trigger them from any process, such as a CRM, a ticketing system, or a nightly batch run. The interface is OpenAPI-compliant, so you can generate clients in any common language. Authentication and permissions follow the same rules as in the interface, and every API call lands in the same complete audit trail. The full OpenAPI spec lives at api.ai-gent.ch. The practical difference: the API is not limited to a subset but covers the platform's full functionality.