Agents that get work done. With permissions you define.
We develop AI agents that handle a clearly defined area of work on their own and, through secure interfaces, access only the systems you allow.
Permissions are what matter: a Finance Agent pre-checks invoices but never triggers a payment itself. An HR Agent structures applications but makes no personnel decisions. Those stay with responsible people.
Sales Agent
- Handles
- Researching and qualifying leads, keeping the CRM up to date.
- Works with
- CRM, email, public company data
- Limit
- Does not send quotes without approval.
Finance Agent
- Handles
- Pre-checking invoices, spotting discrepancies, producing reports.
- Works with
- ERP, accounting, document archive
- Limit
- Never triggers a payment itself.
Service Agent
- Handles
- Categorising customer requests and handling standard cases.
- Works with
- Ticketing system, email, knowledge base
- Limit
- Special cases and complaints go to a person.
HR Agent
- Handles
- Structuring applications, preparing onboarding.
- Works with
- Application inbox, templates, calendar
- Limit
- Makes no personnel decisions. Those stay with people.
Operations Agent
- Handles
- Coordinating projects, tracking tasks, flagging bottlenecks.
- Works with
- Project planning, calendar, documents
- Limit
- Changes budgets and deadlines only after approval.
Research Agent
- Handles
- Researching and summarising markets, competitors and sources.
- Works with
- Approved sources, documents, databases
- Limit
- Cites its sources, decides nothing.
Management Agent
- Handles
- Monitoring KPIs and preparing the basis for decisions.
- Works with
- Reports, management accounting, ERP analyses
- Limit
- Prepares decisions. People make them.
Every step is logged. An agent can be switched off at any time.
