Agents
Set up AI agents that run on event or scheduled triggers to automate repetitive work: follow-ups, lead scoring, status updates, notifications, and reports.
Agents automate the work your team repeats by hand in Reonic. An agent runs on a trigger, does the job you described, and runs again the next time the trigger fires. Think of it as the AI chat working unattended, on a schedule or in response to events in your workspace.
What agents are
An agent is a trigger plus instructions in plain language. The trigger decides when it runs. The instructions describe what it should do, in the same words you would use with a colleague. No formulas, no workflow builder.
Create an agent
- Create a new agent and give it a name
- Choose the trigger: an event or a schedule
- Write the instructions in plain language
- Run it once manually to check the result, then activate it
Event triggers
Event triggers fire when something happens in Reonic: a new request is created, a project changes status, or an offer is signed. Use them for everything that must happen every single time a certain action occurs.
Scheduled triggers
Scheduled triggers run on a clock, for example daily at 9:00 or weekly. Use them for recurring checks and digests: summaries, idle-offer sweeps, anything you would otherwise do every morning by hand.
What agents can do
- Create and update tasks
- Assign colleagues to requests, offers, and projects
- Update records and add notes
- Summarise data into digests and reports
- Notify the right colleague when something needs attention
Example agents
Daily pipeline digest. Runs every day at 9:00. Summarises all offers by status, reports how much volume was newly created, how much sits in the whole pipeline, and how much was signed. It can also flag outliers, for example a few large commercial projects that dominate the numbers.
Route requests by postal code. Runs when a request is created. If the postal code starts with a certain digit, the agent assigns the colleague responsible for that region.
Bring the heat-pump specialist in early. Runs when a request is created. If the notes mention a heat pump, the agent assigns your heat-pump colleague, so the right person is on the project from the first minute.
Kick off the installation on a won deal. Runs when a project reaches the won status. The agent creates the tasks for the back office: book the installation team and issue the invoice, matching whatever your process needs at that point.
Idle-offer follow-up. Runs on a schedule. If an offer sits in the same status for longer than a set time or was never worked on, the agent creates a note and notifies the responsible colleague.
Monitor and edit agents
Open an agent to see its runs and what each run produced. You can run any agent manually at any time, which is the easiest way to test new instructions before activating the trigger. To change what an agent does, edit its instructions; the next run follows the new version.
Start with one repetitive task
Pick the one thing your team does identically many times a week, for example routing new requests, and build that agent first. Run it manually a few times, check the output, then activate it. Add the next agent once the first one has earned your trust.
Things to know
- Agents act on your live workspace data once activated. Test with a manual run before you switch a trigger on.
- Precise instructions beat long ones. Name the condition, the action, and the colleague, and the agent does the rest.
- If you do not see Agents in your workspace yet, contact your Reonic account manager.
Need help?
- Step-by-step questions about this flow → contact your Reonic account manager.
- Feature requests / something missing → drop a note to your account manager; we triage feature feedback continuously.
- Bug reports → include the agent's instructions and the run output in your support message.
Last updated on