Frequently asked questions

An AI agent is software that can understand a request, decide what needs to happen based on instructions you define, use connected tools to take action, and report the result — optionally asking a person to approve before it proceeds.

A general chat assistant mainly answers questions in a conversation. An AI agent built on MazXai is connected to your business systems and can take real actions inside them — such as updating a record, preparing a document or sending a reply — following the rules and approval steps you configure.

No. The AI Agent Builder is designed for business users and uses guided steps, plain-language instructions and visual workflows. Custom integrations may involve technical configuration, which our team can assist with.

Storage location details are documented on the Data Privacy page. [DATA STORAGE LOCATION] should be confirmed there before this answer is finalized for your deployment.

MazXai is built with European businesses in mind and aims to support GDPR-aligned data handling. Specific compliance documentation and a Data Processing Agreement should be confirmed with our team — see the Data Privacy page.

Agents can connect with supported integrations, and with other systems through REST APIs and webhooks. See the Integrations page for current availability — some integrations are marked "coming soon" or available as a custom integration.

Yes. Most businesses build several agents over time, each focused on a specific process or department. Plan limits on the number of agents are listed on the Pricing page.

Yes, for actions you configure as automatic. You decide, per workflow, which actions run without review and which require human approval.

Yes. Human approval steps can be added to any workflow so a person confirms high-impact actions — such as sending a payment or a sensitive customer reply — before they happen.

Yes. Alongside the self-service builder, our team can help design, build or configure agents and integrations for processes that need extra support. Book a demo to discuss your use case.

Yes. Every agent can be run in a safe test environment before it is activated against real data and systems.

Every execution is logged, so mistakes can be reviewed and traced back to a specific run. Adding a human approval step for higher-risk actions is the most effective way to prevent an incorrect action from taking effect in the first place.

Yes. The execution history shows what each agent did, when, and the outcome — giving you visibility into every automated action.

Security is a core part of how the platform is built, including access controls, permissions and logging. See the Security page for details, and the Data Privacy page for how data is handled specifically.

Yes. MazXai supports REST API and webhook connections so agents can be linked to internal systems and custom applications.

Pricing depends on the plan and your usage. See the Pricing page for an overview of example packages, or contact our sales team for a quote tailored to your business.

Yes. Our Enterprise plan covers custom limits, custom AI solutions, dedicated support and onboarding assistance. Book a meeting to discuss your requirements.

Yes, custom integrations can be built for systems not listed on the Integrations page. Reach out to discuss what you need connected.

The orchestration fee is a percentage charged on top of your LLM usage that covers running, coordinating and executing agent workflows on the platform. The percentage varies by plan (for example 25% on Starter Standard and 20% on Starter Premium and Teams Standard) and is waived entirely on Teams Premium and Enterprise.