Product, security, integrations, and rollout. If your question is not here, write to us and a person will answer it.
Two AI voice agents for healthcare practices. The Front Desk agent answers every inbound call — booking, rescheduling, intake, and follow-up, around the clock. The Recruitment agent screens applicants end-to-end: it reads resumes, runs a real voice interview, and scores every candidate against your rubric. You can run either one on its own.
A team that includes a practicing clinician and co-founder of Riverside Pediatrics, a working Connecticut practice. The agents were shaped call by call, alongside real patients and real hires.
It confirms date of birth and the phone number already on the chart, then texts a one-time code to that number and waits for the caller to read it back. The code only ever goes to the number on file, never one supplied fresh on the call, and appointment details are never read out before verification passes.
You define what always escalates — clinical questions, an upset caller, anything outside its scope — and the agent transfers with the context it already gathered. Outside your hours it takes a full message and flags it for the morning.
Calls are answered live at night, on weekends, and on holidays. Anything the agent cannot close it takes a full message for and flags for the morning, so the first person in knows what happened overnight.
Yes. You give us the role, the must-have skills, and the screening criteria, and we build a structured rubric from them. You review and approve it before a single call goes out, and you can change it between roles or between rounds.
Consent is captured before the call begins, and candidates can opt out and request a human interviewer at any point. Opting out does not remove them from the pipeline — the request is passed to your team.
No. It screens, interviews, and scores against your rubric so every candidate gets the same first look. Who advances is your team’s call.
No one, by default. Data stays inside your tenant — it's never used to train a shared model, and is only shared with the subprocessors necessary to run the service.
Most likely. Munchkins reads and writes to major EHR, scheduling, and ATS platforms out of the box. If we don't support yours yet, we'll scope a connector before you sign anything.
No. Both agents run real, natural voice conversations by phone — callers and candidates talk normally, no menus, no forms, no keypad presses.
The agent picks up before the second ring, every time — no hold music and no phone tree. It handles the reason for the call directly: booking, rescheduling, cancellations, intake questions, directions, hours, and insurance questions you have given it answers for. Callers are told they are speaking with an AI from the first sentence.
Yes. Bookings, reschedules, and cancellations write directly to the scheduler you already run, against real availability — visit lengths, provider preferences, and your existing buffers. Nobody re-keys anything afterward, and the caller gets a text confirmation before they hang up.
No. The agent does not triage, interpret symptoms, or give clinical guidance. Clinical questions go to humans, always.
When a candidate clears the resume screen, our system places an outbound call. The agent conducts a structured interview — clinical vocabulary, role-specific questions, real follow-ups when an answer needs clarifying. The full conversation is recorded and transcribed, and you get the audio, transcript, and a scored breakdown for every candidate.
The resume parser pulls out licences and certifications as structured fields — RN, BLS, ACLS, PALS and the rest — with the dates listed. Anything a role requires but the candidate has not listed is surfaced as a gap for your team to check.
Clinical and support roles across a practice — nursing, medical assistants, techs, and front-office staff. The interview is built from the rubric you approve, so the questions match the role rather than a generic template.
Yes. Every interview, resume, and score is encrypted, access-controlled, and retained under the same safeguards as your clinical PHI.
Use the Data Request form and we will verify the request and respond within 30 days. Candidates route requests through the practice they applied to, and we assist that practice in fulfilling them.
A short discovery call, then a pilot configured for one line or one role. Most practices are live within two weeks.
A short call with the people who built it — no script, no sales engineer relay.