Questions and answers.

First Mate provides managed AI employees for businesses. An AI employee has its own Telegram account, email identity, and browser access to your tools.

ChatGPT and Claude are conversational tools. A First Mate AI employee is persistent and operational: it has identity, memory, credentials, and can execute work across systems.

Workflow tools require fixed step-by-step logic. An AI employee can handle open-ended instructions and adapt when the exact path changes.

It can handle browser-based knowledge work such as research, outreach, CRM updates, scheduling, reporting, and content drafting.

It should not be used for high-stakes autonomous decisions such as hiring decisions, legal commitments, or financial approvals without human review.

Setup starts after a 30-minute onboarding call. We configure the employee together, then you grant tool access as you would for a new teammate.

Usually not. The employee works through the browser, so API integrations are not required for most workflows.

Use your normal onboarding process: create accounts, set permissions, and grant access to relevant workspaces.

At the start, the AI employee asks structured questions about your workflows, priorities, and communication style to learn your operating context.

Yes. We offer a seven-day money-back guarantee.

Send tasks via Telegram, Slack, or email. Natural-language instructions are enough for most work.

Yes. It can process tasks asynchronously and report progress or results in your preferred channel.

By default, one verified manager controls instructions. Trusted contacts can be added when needed.

Yes. AI identity disclosure is default behavior in external communication.

Persistent data such as memory, files, and logs is stored in the EU. Some AI inference may involve US processors under SCC-based agreements.

No. Customer data is not used for model training.

You can export data before termination. Service data is permanently deleted within 30 days after termination unless law requires otherwise.

Each AI employee accepts instructions from one verified manager by default. Unauthorized instruction attempts are blocked and logged.

Yes. Billing is monthly and can be canceled without long-term commitment.

Thinking Credits measure AI compute usage. Lightweight tasks use fewer credits and reasoning-heavy tasks use more.

The employee pauses and notifies you. You can top up manually or enable auto top-up.

Different tasks require different models and runtime resources. The platform routes work to balance quality and cost.

We use frontier Claude models including Claude Opus 4.6, Claude Sonnet 4.6, and Claude Haiku 4.5. The platform selects the best model per task automatically.

Each employee runs a sandboxed browser in an isolated container and interacts with web tools like a human operator.

First Mate is built on open-source agent technology with proprietary management, identity, billing, and security layers.

Enterprise customers can deploy in their own infrastructure. Contact us for details.