company logo

Help Center

Go to Aesthetix CRM
Terms & ConditionsPrivacy Policy
All collectionsAI EmployeeAgent StudioManaged Agents: Real Business Use Cases for Medical Practices

Managed Agents: Real Business Use Cases for Medical Practices

Thirty practical Managed Agent ideas for med spas, dermatology and plastic surgery practices, each with a scenario, prerequisites and a customizable sample prompt for lead, marketing, retention and operations work.

Managed Agents in Aesthetix CRM are AI assistants that complete multi-step business tasks from natural-language instructions. This guide gives thirty ways a medical aesthetics practice can use them, each with a scenario, what you need in place and a sample prompt you can paste into the builder at AI Agents > Agent Studio > Managed Agents. If you are new, start with "Agent Studio Overview: Managed Agents, Flow Agents, Conversational AI and Ask AI" and "Managed Agents: Building an Agent with the Guided Builder and Instructions".

Protected health information (PHI): AI features in Aesthetix CRM are not approved for processing PHI. Do not put PHI into agent instructions, prompts, files, memory or knowledge. Every example here is designed around business data such as lead source, consult interest, appointment times, tags and pipeline stage. Do not ask an agent to read treatment notes, diagnoses, medical history, photos, intake or consent forms, or patient surveys that contain health details. Where an idea could drift toward PHI, the caution is called out.

Calendars: AI appointment booking works only with native Aesthetix CRM calendars, not EMR calendars. Appointment-based ideas below (consult prep, no-show recovery, review requests) assume the appointment lives on a native calendar.

How to Read These Examples

Each use case lists:

  • What it does: the outcome.

  • Practice scenario: a realistic situation.

  • Prerequisites: the trigger, apps, Knowledge Base, tags or workflow you should have ready.

  • Sample prompt: a starting point. Replace the bracketed items with your own details, then refine in the builder.

Habits that make every agent safer:

  • Tell the agent never to invent prices, offers, availability or clinical claims, and to say "unknown" when it cannot verify something.

  • Require approval before anything is sent to a patient or published.

  • Add stop rules: stop when the contact replies, books, or opts out.

  • Check for existing records or prior messages to avoid duplicates.

  • Test with the Test Agent button and review results in Agent Logs before publishing. See "Managed Agents: Testing, Publishing, Unpublishing and the Activity Feed".

Lead Management Use Cases

1. Referral Partner Finder

What it does: finds local businesses that match your ideal partner profile and organizes them for outreach.

Practice scenario: a med spa wants to partner with bridal studios, gyms and wellness boutiques for cross-referrals. The agent builds a researched list so the team does not search by hand.

Prerequisites: a supported prospecting or research connector, your partner criteria, and permission to create CRM contacts or write to a connected spreadsheet.

Sample prompt: Create a Referral Partner Finder agent for [Practice Name]. When I ask, find up to [Number] local businesses matching [Business Types] in [City]. Collect verified business names, websites, locations and public contact details. Remove duplicates, mark missing details as unknown and explain why each business fits. Save approved results to our connected Google Sheet, or create CRM contacts only after checking for existing records. Do not start outreach without my approval.

2. New Lead Research and Scoring

What it does: evaluates incoming leads and gives your team a priority.

Practice scenario: consult requests arrive from forms with different budgets and goals. The agent scores each by your criteria so the front desk knows who to call first.

Prerequisites: a lead form, your scoring criteria, CRM permissions. Use only information the lead typed into the form. Do not collect or score on medical history.

Sample prompt: Create a Lead Scoring agent for [Practice Name]. Run when [Lead Form Name] is submitted. Review the submitted service interest, timeline, budget range and lead source. Score 0 to 100 using these weighted criteria: [Criteria and Weights], and classify High, Medium or Low. Explain the score with evidence, add the score and recommended next action to the contact record, apply a priority tag, and create a call task for High leads. Do not use or request medical history. If something cannot be verified, mark it unknown.

3. New Consult Opportunity Brief

What it does: creates a short brief when a new opportunity enters your consult pipeline.

Practice scenario: the patient coordinator wants context before the first call. The agent summarizes the lead's source, stated interest, past messages and open tasks.

Prerequisites: an Opportunities pipeline, the Opportunity Created trigger (see "Managed Agents: Triggers Reference (CRM, Appointment, Scheduled, Webhook) and Follow-Ups"), contact and opportunity access, task permissions.

Sample prompt: Create a Consult Brief agent that runs when an opportunity is created in [Pipeline Name]. Review the opportunity, contact details, notes, tags and message history. Write a brief with service interest, source, stage, previous interactions and suggested next steps, separating confirmed facts from suggestions. Add it as a CRM note and create a follow-up task for the assigned coordinator if none exists. Do not change value, stage or owner. If key information is missing, list questions the coordinator should ask.

4. New Lead Welcome

What it does: welcomes a new lead and points them to the right next step.

Practice scenario: a lead requests information about injectables through the website. The agent identifies the interest and starts the matching welcome workflow.

Prerequisites: a Form Submitted trigger, a Knowledge Base with approved service information, tags, a welcome workflow.

Sample prompt: Create a New Lead Welcome agent for [Practice Name]. Run when [Lead Form Name] is submitted. Identify the service of interest, then use [Knowledge Base] to write a short, friendly message explaining what happens next. Do not invent pricing, availability or results. Add a New Lead tag, record the interest category and trigger the welcome workflow. Skip contacts who already received a welcome and stop if the contact opts out.

5. Consult Preparation Assistant

What it does: prepares a briefing before a scheduled consultation.

Practice scenario: the provider has back-to-back consults. An agent prepares a one-page business-side brief (lead source, services requested on the booking form, prior conversations with the front desk) before each one.

Prerequisites: a native calendar, an appointment trigger, CRM access, a scheduled follow-up, and optionally a Google Docs connector. See the Appointment Triggers section of "Managed Agents: Triggers Reference (CRM, Appointment, Scheduled, Webhook) and Follow-Ups". Keep clinical notes out of the brief.

Sample prompt: Create a Consult Prep agent for [Practice Name]. Run when an appointment is booked on [Calendar Name] and prepare the brief [Time] before the appointment using a scheduled follow-up. Review contact details, earlier front-desk notes, opportunity status and the services requested at booking. Write a concise brief with the consult goal, background, prior interactions and suggested questions. Skip it if the appointment is canceled and use the new time if rescheduled. Never invent patient needs and never include medical history.

6. No-Show Recovery

What it does: follows up after a missed appointment.

Practice scenario: a consult is marked No Show. The agent prepares a kind note with your rebooking link and stops once the person rebooks.

Prerequisites: the Appointment Status Changed trigger on a native calendar, a rescheduling link, tags and a messaging workflow.

Sample prompt: Create a No-Show Recovery agent for [Practice Name]. Run when an appointment on [Calendar Name] changes to No Show. Prepare a polite follow-up that does not assume why they missed it, include [Booking Link], and trigger [No-Show Follow-Up Workflow] to send it. Add a No Show Follow-Up tag and create a task if needed. Do not message again if they rebooked or opted out, and stop when a new appointment is booked.

7. Automated Lead Follow-Up

What it does: runs a consistent follow-up sequence for leads who have not taken the next step.

Practice scenario: a lead received a treatment-plan estimate but has not replied. The agent follows up on a schedule and stops when they respond or book.

Prerequisites: the Tag Changed trigger, CRM history, an approved messaging workflow, scheduled follow-ups. See the follow-up section of "Managed Agents: Triggers Reference (CRM, Appointment, Scheduled, Webhook) and Follow-Ups".

Sample prompt: Create a Lead Follow-Up agent for [Practice Name]. Start when the [Follow-Up Required] tag is added to an active lead. Review prior interactions and opportunity stage, then prepare an initial message, a second after [Number] days and a final after [Number] more days using our approved messaging workflow. Before each message check whether the lead replied, booked, became a patient or opted out, and stop if so. Record activity in the CRM and do not repeat earlier content.

8. Inactive Lead Re-Engagement

What it does: reintroduces your practice to older leads who may still be interested.

Practice scenario: hundreds of leads asked about laser treatments last year and never booked. The agent finds eligible ones and prepares outreach based on their original interest.

Prerequisites: clear inactivity criteria, a scheduled trigger, tags and a re-engagement workflow. Only contact people who have not opted out.

Sample prompt: Create an Inactive Lead Re-Engagement agent for [Practice Name]. Run weekly and review leads matching [Criteria], such as no engagement for [Number] days. Exclude patients, contacts in active conversations, unsubscribed contacts and anyone already in a re-engagement campaign. Group by last known interest, prepare outreach using current offers from [Knowledge Base], add re-engagement tags and start the approved workflow. Record why each lead was chosen, prevent duplicate enrollment and stop on a reply or opt-out.

9. Lost Opportunity Recovery

What it does: reviews lost opportunities and suggests a recovery action when there is a real reason to reconnect.

Practice scenario: some consult leads declined for timing or budget. The agent notes the reason and schedules a later check-in.

Prerequisites: the Opportunity Status Changed trigger, pipeline access, CRM notes, an approved recovery workflow.

Sample prompt: Create a Lost Opportunity Recovery agent for [Practice Name]. Run when an opportunity in [Pipeline Name] changes to Lost. Review notes, the recorded loss reason, communications and contact preferences. Categorize as Timing, Budget, Offer Fit, Competitor or Other only when the evidence supports it, then recommend a recovery action. If eligible, create a future follow-up task or enroll in [Recovery Workflow]. Use only verified offers. Do not reach out if the contact opted out or declined further contact.

10. Daily Pipeline Digest

What it does: summarizes pipeline activity and highlights deals that need attention.

Practice scenario: the practice manager wants a morning view of high-value treatment-plan opportunities and stalled follow-ups.

Prerequisites: opportunity read access, a scheduled trigger, and optionally a Google Docs connector.

Sample prompt: Create a Daily Pipeline Digest agent for [Practice Name]. Run every weekday at [Time] in our account timezone. Review opportunities in [Pipeline Names], their stages, values, recent activity and upcoming tasks. Report total open pipeline value, high-value opportunities, opportunities with no activity for [Number] days and deals needing follow-up. Highlight missing information and do not treat an empty activity log as proof nobody spoke to the lead elsewhere. Save the report to [Approved Destination]. Do not change records.

11. CRM Data Quality Assistant

What it does: finds incomplete or inconsistent contact records.

Practice scenario: contacts have missing names, bad phone formats and inconsistent tags. The agent prepares a cleanup report.

Prerequisites: CRM read access, your data rules, a scheduled trigger, and permission for approved updates or tasks.

Sample prompt: Create a CRM Data Quality agent for [Practice Name]. Run every [Frequency] over [Defined Segment]. Find contacts with missing required fields, inconsistent tags or incomplete names. Report the contact, issue and suggested fix. Apply only safe updates that follow our written rules. Flag suspected duplicates for review instead of merging or deleting. Never invent missing information. Create review tasks for unresolved items.

Marketing and Content Use Cases

12. Social Media Posting

What it does: drafts branded posts and schedules approved ones.

Practice scenario: the practice wants three posts a week about seasonal skin care education and events, with no before-and-after photos created or reused without documented patient consent.

Prerequisites: a marketing Knowledge Base, brand guidelines, connected social accounts, and optionally image generation. See the Social Planner guides for publishing.

Sample prompt: Create a Social Media Posting agent for [Practice Name]. Prepare [Number] posts per week for [Platforms] using [Marketing Knowledge Base] and our brand guidelines on [Content Pillars]. For each post write a caption, hashtags, a call to action and an image brief. Never invent offers, prices or result claims, and never reference or generate patient photos. Present drafts for approval and schedule only after approval.

13. Email Campaign Manager

What it does: prepares email campaigns for chosen contact segments.

Practice scenario: a launch of a new membership program for existing patients. The agent drafts the campaign for your review.

Prerequisites: a marketing Knowledge Base, CRM segments, email sending configured, an approved campaign workflow.

Sample prompt: Create an Email Campaign Manager agent for [Practice Name]. Build campaigns for contacts matching [Audience Criteria] using [Marketing Knowledge Base]. For each, write an objective, audience description, two subject lines, preview text, body, call to action and follow-up recommendation. Personalize using CRM data without assumptions about preferences or treatments received. Exclude unsubscribed contacts. Submit content and segment for review before activating the workflow.

14. Weekly Creative Studio

What it does: produces a batch of marketing creative each week.

Practice scenario: weekly promotional graphics and short video ideas for upcoming events or seasonal services.

Prerequisites: a marketing Knowledge Base, a scheduled trigger, image or video generation enabled, brand assets.

Sample prompt: Create a Weekly Creative Studio agent for [Practice Name]. Run every [Day] at [Time]. From [Marketing Knowledge Base], review themes, approved offers and brand guidelines. Generate [Number] concepts with image ideas, short-video concepts, headlines and captions in [Aspect Ratios]. Save generated media to Media Storage with a clear summary. Support every claim with approved information, do not depict patients, and do not publish without review.

15. Content Repurposing Assistant

What it does: turns one approved piece of content into several formats.

Practice scenario: a monthly patient education webinar transcript becomes social posts, an email and short-video scripts.

Prerequisites: the approved source document, a brand Knowledge Base, optionally a Google Docs connector.

Sample prompt: Create a Content Repurposing agent for [Practice Name]. When our team provides an approved article or webinar transcript, extract the main ideas and takeaways, then write [Number] social posts, [Number] captions, one newsletter draft and [Number] short-video scripts in our brand voice from [Marketing Knowledge Base]. Preserve meaning, avoid unsupported claims, and save drafts to Google Docs for review.

16. Competitor Watch

What it does: tracks public competitor information and flags meaningful changes.

Practice scenario: monitor nearby practices' published service menus and promotions weekly.

Prerequisites: a scheduled trigger, a connected research tool, a competitor website list, optionally Google Sheets or Docs. If you use the separate Computer capability for browsing, run the agent manually or in chat, because Computer does not run during triggers. See "Managed Agents: Using Computer (Browser, Files and Terminal)".

Sample prompt: Create a Competitor Watch agent for [Practice Name]. Every [Frequency], review these approved competitor websites: [Website List]. Use our connected research tool to retrieve published services, pricing and announcements. Compare with previous results in [Sheet or Document], highlight real changes with source URLs, and separate confirmed changes from possible ones. Do not access restricted information.

17. Marketing Performance Report

What it does: reviews ad performance and writes a report.

Practice scenario: a monthly summary of Facebook and Google ad results for the practice's campaigns.

Prerequisites: connected ad accounts, Facebook Ads Reporting or Google Ads Reporting permission, a scheduled trigger. The supported reporting operations are listed in "Managed Agents: Custom Skills, Files and Memory".

Sample prompt: Create a Marketing Performance Report agent for [Practice Name]. Run every [Frequency] and retrieve Facebook and Google Ads reporting for [Reporting Period]. Analyze spend, impressions, clicks and conversions, compare with the previous period, and name the strongest and weakest campaigns with clear criteria. Report key metrics, changes, possible issues and recommendations. Do not invent missing metrics and do not change campaigns or budgets.

18. Lead Source Quality Report

What it does: shows which lead sources produce booked consults, not just clicks.

Practice scenario: compare Instagram, Google, referrals and website forms by qualified leads and won treatment plans.

Prerequisites: consistent lead source tracking, CRM reporting access, qualification criteria, a scheduled trigger.

Sample prompt: Create a Lead Source Quality agent for [Practice Name]. Run monthly over [Reporting Period] using lead source fields, opportunity statuses and qualification tags. Group by source, count leads, qualified leads and won opportunities where data supports it, and list unattributed records as Unknown. Describe patterns without assuming causation. Do not change attribution, statuses or campaigns.

19. Review Request Coordinator

What it does: starts approved review requests after eligible visits.

Practice scenario: after a completed visit, the agent starts your review workflow once per patient. It uses only appointment status or a tag, not treatment details.

Prerequisites: the Appointment Status Changed or Tag Changed trigger, completed-visit criteria, reputation settings, a review request workflow. Appointment triggers apply to native calendars.

Sample prompt: Create a Review Request agent for [Practice Name]. Run when an appointment on [Service Calendar] changes to Showed or the [Visit Completed] tag is applied. Confirm eligibility and that they were not asked recently, then start [Review Request Workflow] through an approved channel. Record that the request started. Do not fake reviews, offer incentives for positive ratings, or ask only happy patients for public reviews. Respect contact preferences and message limits. Do not mention treatment details in the message.

20. Experience Feedback Summary

What it does: summarizes non-clinical feedback themes such as wait times, front-desk friendliness and booking ease.

Practice scenario: a short post-visit experience survey asks only about scheduling, communication and facility. The agent groups themes and creates follow-up tasks.

Prerequisites: a survey that does not ask for health or treatment information, the Survey Submitted trigger, CRM access, task permissions. Caution: patient feedback often drifts into clinical detail such as symptoms or results. Do not point this agent at surveys that may contain PHI. If a response mentions health information, a staff member should handle it, not the agent.

Sample prompt: Create an Experience Feedback agent for [Practice Name]. Run when [Survey Name] is submitted. Identify themes and sentiment about scheduling, staff communication and facility only. Add a short factual summary to the contact and apply a feedback tag. For complaints, create an internal task with the stated concern and a suggested next step. If the response mentions symptoms, treatment outcomes or other health information, do not summarize it and create a task for a staff member instead. Never promise refunds or exceptions.

Patient Support, Onboarding and Retention Use Cases

21. New Patient Onboarding Coordinator

What it does: coordinates tasks and messages when a lead becomes a patient.

Practice scenario: when a treatment-plan opportunity is marked Won, the agent sends your welcome workflow and creates the front-desk setup tasks.

Prerequisites: the Opportunity Status Changed trigger, an onboarding Knowledge Base, tags, tasks, an onboarding workflow.

Sample prompt: Create a New Patient Onboarding agent for [Practice Name]. Run when an opportunity in [Pipeline] changes to Won. Review contact details and our [Onboarding Knowledge Base], prepare a welcome summary and list the required onboarding steps. Add Patient and Onboarding tags, trigger [Onboarding Workflow] and create setup tasks with clear descriptions. Check whether onboarding already started. Flag missing information for the coordinator instead of guessing. Do not request or record medical information.

22. Knowledge Base Front-Desk Assistant

What it does: answers common questions from your approved documentation.

Practice scenario: website visitors ask about hours, parking, financing options and what to expect at a first consult.

Prerequisites: an indexed Knowledge Base, a supported chat channel, CRM access, an escalation workflow. For a chat bot that answers customers all day, also see Conversational AI.

Sample prompt: Create a Front-Desk Assistant for [Practice Name] on [Chat Channel]. Use [Support Knowledge Base] as the only source for policies, hours and general information. Answer concisely and ask a clarifying question when needed. Do not invent services, prices or policies, and do not give medical advice or discuss individual health concerns. If the Knowledge Base has no reliable answer, say a team member will follow up and create a task or trigger the escalation workflow.

23. Inbound Request Triage

What it does: classifies incoming requests and routes them to the right team member.

Practice scenario: requests arrive through a contact form about billing, rescheduling, product questions and general inquiries.

Prerequisites: a request form, urgency criteria, tags, tasks, routing rules. For advanced routing through a workflow, see "Managed Agents: Invoke from Workflows and Deploy to API". Instruct the form not to collect health details.

Sample prompt: Create a Triage agent for [Practice Name]. Run when [Request Form] is submitted. Classify the request into [Categories], judge urgency by our criteria and summarize the issue and recommended next action. Add the right tag and create a task using our routing rules. Mark billing disputes and complaints for priority human review. Never say an issue is resolved unless an action confirms it, and avoid duplicate tasks. If the message contains health information, route it to a staff member without summarizing it.

24. Membership Community Moderation

What it does: helps review member questions and flag guideline violations in a general wellness or membership community.

Practice scenario: a community for your skin-care membership program where members ask product and routine questions.

Prerequisites: community guidelines in a Knowledge Base and a supported way to pass content to the agent. Automatic access to community activity needs a separately verified trigger or integration. Caution: members may post health details. The agent should not engage with them; route those posts to a staff member.

Sample prompt: Create a Community Moderation agent for [Community Name]. Use [Community Guidelines Knowledge Base] to answer common questions and review posts for spam, harassment or prohibited promotions. Explain which guideline may apply and prepare a recommendation. Flag uncertain, sensitive or health-related posts for human review. Do not remove posts, ban members or send disciplinary messages unless explicitly configured and approved.

25. Patient Education Course Follow-Up

What it does: reviews course enrollment and progress and prepares encouragement.

Practice scenario: a pre-treatment education course for general topics such as skincare routines. The agent finds learners who have not finished the key lesson.

Prerequisites: a course, course permissions, progress criteria, a scheduled trigger, an optional follow-up workflow.

Sample prompt: Create a Course Progress agent for [Practice Name]. Run weekly and use the supported Courses capabilities for [Course Name]. Find learners enrolled without completing [Milestone] after [Number] days, summarize their progress, write a friendly encouragement using [Course Support Knowledge Base] and trigger the follow-up workflow where appropriate. Respect contact preferences, avoid repeat reminders, and do not change enrollments without authorization.

26. Membership Renewal Coordinator

What it does: prepares account managers or front-desk staff for upcoming membership renewals.

Practice scenario: members on a monthly skin-care membership approach renewal. The agent prepares a renewal brief and a draft message for review.

Prerequisites: renewal dates and membership details in accessible data, a date-based workflow that applies a renewal tag, CRM and task permissions.

Sample prompt: Create a Membership Renewal agent for [Practice Name]. Start when the [Renewal Due] tag is applied. Review the contact record, membership details, renewal date, open issues and notes. Prepare a brief with verified details, concerns and suggested questions, create a preparation task, and draft a message for review. Do not invent terms, discounts or dates, do not modify memberships or charge anyone, and flag conflicting information for review.

Operations and Team Productivity Use Cases

27. Hiring Process Coordinator

What it does: organizes applicant information and creates tasks for the recruiter.

Practice scenario: applications for a medical aesthetician role arrive through a form.

Prerequisites: an application form, role requirements, a tracking destination, optionally a calendar or project connector.

Sample prompt: Create a Hiring Coordinator for [Practice Name]. Run when [Application Form Name] is submitted. Summarize job-related qualifications, license or certification details the applicant provided, relevant experience and missing information against [Job Role] requirements. Save the summary to [Tracking Destination] and create a recruiter review task. Do not make hiring decisions, infer protected personal characteristics or reject candidates automatically. Flag incomplete applications.

28. Meeting Notes and Action Items

What it does: turns a meeting summary into tasks.

Practice scenario: the weekly marketing and operations meeting produces action items that need owners and due dates.

Prerequisites: a meeting transcript or a connected meeting-notes app, and a task management connector. Use non-clinical staff meetings only; do not use case reviews that discuss patients.

Sample prompt: Create a Meeting Action Item agent for [Practice Name]. When I provide an approved meeting transcript or summary, identify objectives, decisions, open questions and explicitly agreed action items with owner and due date when stated. Do not invent commitments. Create tasks in [Task Tool], mark items missing an owner or deadline as needing clarification, include a source reference, and avoid duplicates.

29. Event and Launch Kickoff Coordinator

What it does: turns a confirmed event or product launch into a kickoff checklist.

Practice scenario: the practice schedules an injector training day or an open house.

Prerequisites: a trigger such as Opportunity Status Changed or a manual request, documented event scope, a project Knowledge Base, a supported project connector.

Sample prompt: Create an Event Kickoff agent for [Practice Name]. Run when [Trigger]. Review the approved scope, date, venue and [Event Planning Knowledge Base]. Produce a kickoff checklist covering materials, vendors, promotion, guest communications and staffing, create a project in [Project Tool] with tasks from our template, and prepare an agenda for the kickoff meeting. Do not invent dates, budgets or vendor commitments, flag missing details and avoid duplicate projects.

30. Internal SOP and FAQ Creator

What it does: turns approved process notes into draft standard operating procedures and FAQs.

Practice scenario: documenting front-desk steps for new consult bookings and rescheduling.

Prerequisites: an internal Knowledge Base or approved source documents, a connected document tool, a defined approval process.

Sample prompt: Create an Internal Documentation Assistant for [Practice Name]. When an authorized team member provides a process name and approved sources, use [Internal Knowledge Base] to draft a standard operating procedure (purpose, prerequisites, roles, steps, expected outcomes, exceptions) or an FAQ based only on verified content. List missing or conflicting instructions as questions, save the draft to Google Docs for human review, and do not publish or alter the Knowledge Base. Do not include HIPAA-regulated patient details in any document.

Setting Up a Use Case

Every example is built from the same three pieces. See the related guides for each.

  1. Trigger: what starts the agent. Options include CRM events (contact or opportunity), appointments, scheduled runs, inbound webhooks, workflow actions and API calls. See "Managed Agents: Triggers Reference (CRM, Appointment, Scheduled, Webhook) and Follow-Ups" and "Managed Agents: Invoke from Workflows and Deploy to API".

  2. Apps and tools: what the agent can use, such as Google Sheets, task tools or research tools. See "Managed Agents: Apps, MCP Connectors and Custom API Calls".

  3. Instructions: the prompt. Start from a sample prompt above. Reusable processes belong in Custom Skills and recurring tasks benefit from Memory. See "Managed Agents: Custom Skills, Files and Memory". Larger jobs can use specialist helpers. See "Managed Agents: Subagents".

Steps:

  1. Open AI Agents > Agent Studio > Managed Agents and click + Create Agent.

  2. Paste a sample prompt and replace the bracketed items.

  3. Connect the apps and add the trigger.

  4. Test with realistic, fictional data (never a real patient).

  5. Review in Agent Logs, adjust, then publish.

Frequently Asked Questions

Can I use these prompts as written?

Use them as a starting point. Replace the bracketed items with your own pipelines, forms, calendars, tags, workflows and Knowledge Bases, then test and refine.

Do I need every prerequisite before I try one?

You need the trigger, access and workflows that the example relies on. Missing pieces, such as an approved messaging workflow, should be built first.

Which examples are safest to start with?

Internal ones that do not message patients: Daily Pipeline Digest, CRM Data Quality, Lead Source Quality, Meeting Action Items and SOP creation.

Can an agent book appointments for me?

Yes, but only on native Aesthetix CRM calendars. AI appointment booking does not work with EMR calendars.

Can I send patient photos, charts or intake forms to any of these agents?

No. AI features are not approved for PHI. Keep patient health information out of agent prompts, files, memory, knowledge and connected tools.

How do I stop an agent from messaging someone twice?

Add stop rules to the prompt (replied, booked, opted out) and have the agent check for earlier messages and tags before sending. Test the logic before publishing.

Why does my Computer-based idea not run on a trigger?

Computer works in chat and manual runs only. Use a connected research tool or app instead if the agent must run on a schedule or trigger.

Where do I see what an agent did?

Open the agent's Activity tab and Agent Logs. See "Agent Studio: Agent Logs, Metrics, Monitors and AI Usage".

Did this answer your question?
😞
😐
😁