Fred for medical practices.
Answer patients day and night, book and route to your front desk, and protect every record around the clock, an AI agent that captures the visit and never diagnoses, prescribes, or improvises a medical answer.
Answer the patient. Never cross the line.
Patients ask at all hours: what you treat, what you accept, when you can be seen. The practice that answers books the visit. But an AI that improvises a medical answer is a liability no practice can carry. Fred answers what it should, books the appointment, and refuses to diagnose or prescribe, every single time.
What Fred does
- Answers hours, location, services, and what you treat
- Books, confirms, and reschedules appointments
- Routes urgent symptoms to 911 and your care team
- Captures new-patient information securely
When Fred hands off to you
- A medical emergency or worrying symptoms
- Diagnosis, treatment, or prescription requests
- Billing and insurance claims
- Anyone asking to speak with a clinician
What Fred won’t do
- Diagnose, interpret symptoms, or guess what’s wrong
- Prescribe, refill, or adjust medication
- Handle protected health information loosely
Pre-configured for medical practices: primary care, family medicine, clinics, and multi-provider groups; appointments, new-patient intake, and general questions.
The question comes in at 2am. Fred answers.
Talk to Fred.
Server-side guardrails
The guardrails you manage in the admin, enforced before Fred replies.
Every medical practice deployment runs the Healthcare & Med Spa pack. This is the compliance panel you manage in the admin: the industry pack automation, your own custom rules, and the hard limits Fred holds no matter how it’s asked. All of it is enforced in code before a word reaches the visitor. Not a suggestion in a prompt a clever question can talk around.
Automated Guardrail Actions
Pack rules enforced in addition to baseline protections.industry.healthcare
| Rule ID | Action | Severity | Phase | Message | Triggers |
|---|---|---|---|---|---|
| hc-emergency-911 | ESCALATE | S4 | input | This sounds like a medical emergency. Please call 911 immediately. I'm alerting the care team now. | contains: chest pain, can't breathe, difficulty breathing +49 |
| hc-no-diagnosis | ESCALATE | S4 | input | I can't evaluate symptoms or diagnose conditions — that requires a clinical examination. Let me capture your information so a member of our care team can follow up with you. If this feels urgent, please call our office directly… | contains: diagnose, what do I have, is this cancer +33 |
| hc-no-prescription | ESCALATE | S4 | input | Medication decisions require your care provider. Let me capture your information so our clinical team can follow up with you about your prescription needs. | contains: prescribe, prescribe me, write me a prescription +33 |
| hc-no-personalized-advice | ESCALATE | S3 | input | I can share general health information, but personalized medical advice requires your care provider. Let me capture your information so our clinical team can follow up with you. | contains: what should I do about my, advice for my condition, is this normal for me +26 |
| hc-age-gate | CLARIFY | S3 | input | Some cosmetic and medical spa treatments have minimum age requirements. Can you confirm you are 18 or older? | contains: botox, fillers, lip fillers +44 |
| hc-billing-escalate | ESCALATE | S3 | input | Billing and insurance questions are handled by our billing team to ensure accuracy and privacy. Let me capture your information so they can follow up with you directly. | contains: my bill, my statement, insurance claim +48 |
| hc-phi-redact | REDACT | S3 | input | I've removed protected health information to keep your data secure. We only collect what's needed to connect you with the right care team. | contains: lab result, lab results, blood test results +41 |
| hc-off-label-warning | DISCLAIMER | S2 | output | contains: treatment results, procedure results, clinical results +22 | |
| hc-no-referral-inducement | DENY | S4 | output | I can share general information about our services and help you schedule. For referrals to specialists or other providers, your care team can help coordinate based on your clinical needs. | contains: we'll refer you to our, our partner doctor, our affiliated +19 |
| hc-emtala-compliance | DENY | S4 | output | If you're experiencing a medical emergency, come to our emergency department or call 911 right away. Emergency care is provided regardless of insurance or ability to pay. | contains: insurance before emergency, need insurance to be seen, verify insurance before +11 |
| hc-telehealth-boundary | ESCALATE | S3 | input | This chat isn't a telehealth visit and I can't provide clinical assessments here. If you'd like a virtual appointment with a provider, let me help you schedule one through our telehealth service. | contains: telehealth visit, telemedicine visit, virtual visit +21 |
| hc-results-disclaimer | DISCLAIMER | S2 | output | contains: before and after, dramatic results, amazing results +19 | |
| hc-informed-consent-boundary | ESCALATE | S3 | input | Informed consent and detailed risk discussions are an important part of your care that happens directly with your provider. Let me help you schedule a consultation so your care team can walk you through everything. | contains: sign consent, consent form, sign the form +23 |
| hc-no-outcome-guarantees | DENY | S3 | output | Treatment outcomes depend on individual factors that your care team will discuss with you during your consultation. I can share general information about what to expect. | contains: guaranteed to work, guarantee the results, will fix your +17 |
Hard limits
The boundaries Fred holds, no matter how a visitor asks.
- Never diagnoses conditions or interprets symptoms, that needs a clinical exam
- Never prescribes, adjusts, or refills medication
- Routes medical emergencies to 911 and your care team instantly
- Redacts protected health information and never guarantees an outcome
Your custom rules
Layered on top of the pack, unique to your practice. No developer required.
The pack is the starting line, not the ceiling. Every deployment can add its own rules in the admin, block, warn, or just log on any phrase you choose, in either phase, with your own response message. They run right alongside the pack.
- BLOCKoff-topic requests that aren’t your practice
- WARNbefore anyone is quoted a price over chat
- LOGwhen a visitor asks about a specific provider
These industry rules stack on top of the always-on baseline, your region’s privacy and crisis rules (HIPAA-aware), and macro overlays. See how the guardrail layers work
Know which lead to call back first.
Every conversation lands in your CRM with a lead score, a purchase probability, and a plain-language summary. The ready-to-book new patient and the procedure inquiry get called before the general hours question, so your front desk works the highest-value lead first instead of guessing.
Medical practice questions.
Will Fred give medical advice or diagnose?
No. Fred shares general information only. It won’t evaluate symptoms, diagnose conditions, or recommend treatment, those require your clinical team. It captures the details and routes the patient.
What happens in an emergency?
Fred recognizes emergency language, tells the patient to call 911 immediately, and alerts your care team right away rather than trying to triage it over chat.
How does Fred handle private health information?
Fred is built for healthcare: it redacts protected health information, collects only what’s needed to connect the patient with the right team, and routes records and billing requests through proper channels.
Can Fred prescribe or refill medication?
No. Medication decisions require your provider. Fred captures the request and connects the patient with your clinical team.
Answer every patient. Stay in your lane.
Go live in about 20 minutes. Or ask Fred a question right now, it is in the corner of this page.