How to set up custom actions for dermatology patient inta…

Make sure your Chatref agent already has a knowledge base that covers your practice’s intake process, forms, and common patient questions. The agent needs to understand your workflow so it can guide patients naturally. If you haven’t set up your knowledge base, start with the Dermatology Practices guide to train your agent on appointment policies, accepted insurances, and intake instructions.

You’ll also need:

• A list of the intake fields you want to collect – typical dermatology fields include patient full name, date of birth, primary skin concern, current medications, insurance provider, member ID, and preferred appointment time.
• A webhook URL that can receive the collected data. Many practices send the intake payload to a practice management or EHR system via Zapier or a similar connector, or simply email a formatted intake form to the front desk.
• Access to your agent’s Custom Actions settings in the Chatref dashboard.

1. Open the agent settings – In your Chatref dashboard, select the agent you want to configure for intake and go to its Custom Actions section.
2. Create a new action – Click Add a custom action and give it a name like “Dermatology New Patient Intake”. Write a short description that helps the agent recognise when to trigger it, e.g., “Use this when a new patient wants to start treatment or asks how to become a patient.”
3. Define the fields the agent will collect – Add each data field you decided on, marking the essential ones as required. Good field names for dermatology intake:
   • patient_full_name (required)
   • dob (required, the agent will prompt in MM/DD/YYYY format)
   • primary_concern (required – a text field for the main skin issue)
   • history – previous skin conditions or treatments
   • current_medications – list any current meds
   • insurance_provider – free text or select from options
   • insurance_member_id
   • preferred_appointment_date_time
   • photo_consent – yes/no checkbox for telemedicine intake The agent will ask for each field conversationally, using your knowledge base to answer side questions (“Do you accept my plan?”) while it collects details.
4. Set the trigger – Under Action, choose Webhook (or the option for sending data to an external URL). Enter your webhook URL and, if required, basic authentication headers. The action will send the collected fields as a JSON payload when the conversation finishes.
5. Add a confirmation message (optional) – You can specify what the agent says after a successful intake, for example: “Your intake details have been sent to our team. We’ll confirm your appointment soon.”
6. Save the action – The agent will now offer this intake flow whenever a new patient shows interest.

• Test in the playground – Open the live test chat and pose as a new patient. Say, “I’d like to become a new patient for acne treatment.” The agent should ask the defined questions, gather the fields, and then trigger the webhook.
• Inspect the conversation – After a test, go to the Conversation Inbox to see the full transcript. Every collected field appears in the chat log. Also verify that the webhook delivered the data correctly – check your email or integration tool.
• Test edge cases – Try submitting without a required field; the agent must prompt for the missing item. Give an invalid date or a sparse answer to confirm the flow still completes gracefully.

• The agent doesn’t start the intake – Make sure the action’s description covers the phrases your patients actually use (“new patient”, “become a patient”, “start treatment”). You can also add a line to your agent’s main instructions that it should offer the intake whenever a new patient conversation begins.
• Fields aren’t collected completely – Write specific, friendly field descriptions that tell the agent how to ask. For example, “Ask for the patient’s insurance member ID exactly as it appears on the card” or “Ask for the main skin concern in the patient’s own words.”
• Webhook errors – Check the webhook URL and authentication in a tool like webhook.site before linking it to your practice system. Most failures are a typo in the URL or a missing header.
• Data confusion – Confirm the JSON field names in Chatref match what your downstream system expects (e.g., patient_full_name vs full_name). Use the playground to inspect the payload.
• Patient drop-off – Long intake forms can discourage people. Keep the field list as short as possible, and let the agent use a warm conversational tone. Review transcripts to spot questions that cause patients to hesitate, then simplify those steps.

What causes dermatology patient intake automation problems for Dermatology Practices?

Problems often trace back to three areas. First, incomplete or vague custom-action field definitions cause the agent to skip questions or ask them in a jumble. Second, webhook misconfigurations – wrong URLs, missing auth, or payload mismatches – mean the intake data never lands in your system. Third, the agent’s knowledge base might be missing up-to-date intake-related information, so it can’t answer follow‑up patient questions with authority and may conflict with your actual policies. When any of these pieces are off, the automation looks broken to patients and your team.

How do I improve dermatology patient intake automation for Dermatology Practices?

Refine the intake action to ask only what’s essential, and write clear, approachable prompts in the field descriptions. Keep your knowledge base current with your latest intake forms, insurance lists, and practice policies, then re-train the agent any time you change a process. Test the action with realistic conversations – including incomplete or unusual replies – and watch for webhook logs that show repeated failures. Finally, have a team member review the intake transcripts every week and shorten or rephrase any question that patients struggle with. Small, iterative adjustments turn a functional intake into a frictionless one.