Industries · Healthcare
Patient support with bounded autonomy
Members ask about scheduling, coverage, billing, and portal access — high volume, and every answer has to stay inside the lines. Airclou Helpdesk drafts replies grounded in the record, acts only within limits you set, and sends anything clinical — or anything it’s unsure of — to a person.
Free 14-day trial · No credit card required
Built for healthcare
The AI answers the routine, a person owns the rest
Your staff stops fielding every scheduling and coverage question by hand — while the cases that need a person, or a license, always reach one. You set where the line sits.
Replies drafted from the record, flagged when unsure
When a member writes, the draft is already there — grounded in your help center and the live account, with every source cited. Anything the AI is not confident about, and anything clinical, is flagged for a person instead of being sent.
- Account panel fetched from the system of record, never guessed
- Every draft cites its sources — no unattributed claims
- Low-confidence and clinical drafts are never auto-sent
You set exactly what the AI may do alone
Auto-approval envelopes define which reply kinds the AI may send on its own and cap any action by amount. Everything outside those limits parks at an approval gate with the verified details in front of a person — nothing takes effect from a stray toggle.
- Per-action limits by reply kind and amount
- Anything outside the envelope parks for human approval
- Every action logged with who approved it and why
Route clinical, billing, and scheduling to the right team
Plain-language rules send each message to the team that should handle it and escalate the ones that can’t wait. A clinical question reaches licensed staff while a billing query is handled at the front desk — and an urgent case is flagged while there is still time to act.
- Response targets by case type, with breach warnings
- Rules you can read — when, and, then — no consultant
- Urgent cases escalate before a target slips
Everything included
Every way a member reaches you, in one queue
Portal chat, email, and a public help center land in the same prioritized inbox — with the same record context and the same boundaries behind all of them.
Portal live chat
A fast widget on your member portal — AI answers routine questions from your help center instantly, hands off to a person the moment it should.
Knowledgebase
A help center that deflects the routine coverage and scheduling questions — and grounds every AI draft in your own approved content.
Email on your domain
support@yourorg.com in minutes — replies come from your address, threaded correctly, not from ours.
Response SLAs
Response targets by case type, with breach warnings before an urgent case waits too long.
Automation you can read
Plain-language rules — when, and, then — that route by case type, tag by topic, and escalate without a consultant.
Reporting & CSAT
Deflection, resolution times, and satisfaction — measured continuously, so you can see support quality and where a person’s time is best spent.
Common questions
What healthcare teams ask us first
How do you keep the AI away from clinical advice?
You draw the line, and the product holds it. The AI drafts only within the boundaries you set, low-confidence drafts are flagged for review instead of being sent, and plain-language rules route anything clinical to a licensed person — with the context already attached. The AI answers the scheduling, coverage, and billing questions; it does not practice medicine.
Can we limit what the AI does on its own?
Yes — that is the core of it. Auto-approval envelopes define exactly which reply kinds the AI may send alone and cap any action by amount; everything outside those limits parks at an approval gate with the verified details and Approve / Reject buttons. Execution is idempotent, and every action is logged with who approved it.
Where does our data live, and is it secure?
Every connection is encrypted with TLS, data is scoped per tenant so one organization can never see another’s, access is least-privilege, and every AI action is logged. We handle personal data in line with GDPR. We do not yet hold third-party certifications such as SOC 2 or HIPAA attestation — see the Security page for exactly what is in place today, so you can judge the fit against your own requirements.
We run a legacy system today — how do we evaluate?
Run Airclou alongside it. Forward a slice of your member or patient support address, connect your records in read-only terms you control, and judge the drafts against two weeks of your real queue — no cutover, no risk to your current setup.
Get started
Member questions, answered — inside the lines
Connect your support address and your records, import your help center, and judge the drafts on your own queue — while whatever you use today keeps running.
Free 14-day trial · No credit card · Cancel anytime