When most of your queue is healthcare, the challenge isn't variety — it's density. Here's how to put every drug name, dosage, and date on screen without anything touching the call.
Ask interpreters who take Propio calls what the work is like and you hear the same theme: healthcare, most of the day, most days. Medication reviews, symptom intake, appointment scheduling, benefits questions — call shapes you'll know by heart within a month, each one carrying terminology that punishes autopilot. When the call mix is this medical, the tool question isn't about variety — it's about density: how many precise terms per minute can you verify instead of hold. This guide covers running a live caption and translation layer beside that queue, on phone and video alike.
Propio serves a large healthcare client base, and the queue reflects it: clinics, hospital departments, pharmacies, insurers, and care coordinators make up much of the traffic for many language pairs, with government and business calls mixed in. Work spans both over-the-phone and video sessions, often in the same week, and interpreters generally log in for availability windows — which means the day's rhythm is set by the queue, not by you. Some stretches run back-to-back; others go quiet, and the quiet stretches are their own kind of tiring.
The paradox of a medical-heavy queue is that repetition doesn't make it easier. A medication review is a familiar shape, but the payload — drug names, dosages, frequencies, contraindications — is different every time, and it's precisely the part where "close enough" isn't. Familiarity invites autopilot; the terminology forbids it. That tension is the job.
A single clinical call asks you to hold a lot at once: the sentence being spoken, the sentence you're rendering, and the precise items inside both — a dosage in milligrams, a follow-up date, the name of a specialist clinic. Audio-only work strips the visual cues; video restores a face but not a chart. Either way, working memory is the bottleneck, and in medical interpreting the items that overflow first are exactly the ones that matter most. Our deep dive on Spanish–English medical interpreting walks through where those failure points cluster.
Unicaption is a real-time captioning and translation copilot that runs beside your calls — phone or video — without ever being part of them. It reads your microphone or system audio, so your agency's platform never knows it's there:
On a medical-heavy queue, the custom dictionary is the highest-leverage feature you have, because your calls repeat their vocabulary even when they don't repeat their content. Treat it like a garden: seed it before your first session, weed it monthly, and add to it the moment a term surprises you. How custom dictionaries improve terminology accuracy covers the mechanics; the strategy below is queue-specific.
Start with the terms you already know get misheard: near-homophone drug names, regional anatomy terms, the insurance phrases that have no clean equivalent in your target language.
Clinic names, hospital departments, and specialist names recur constantly in scheduling calls and are exactly what generic speech recognition mangles. Ten minutes of entry saves a hundred small corrections.
The term that made you hesitate on today's call will reappear — same clients, same clinical domains. Add it between calls or right after the shift, while you still remember how it was actually pronounced.
The dictionary is yours, not the queue's. If you also take calls elsewhere, the same accumulated vocabulary follows you — see our other agency guides for how the mix changes.
Most of what crosses a medical-heavy queue is protected health information, and the bar for any tool you run beside it should be simple: it must hold nothing. Unicaption holds nothing — sessions are end-to-end encrypted, audio is never stored, transcripts are deleted when the session ends, and session content is never used to train models. It is end-to-end encrypted with nothing retained after the session, and there is no recording, archive, or history to secure because none is created.
Just as plainly: agencies set their own policies on aids and tools, and following yours is part of the professionalism the job runs on. Check before you assume, and if the guidance is unclear, ask — that's not caution for its own sake, it's the same discipline you apply to everything else on a healthcare call.
Your agency's own policy is the authority on that, and checking it is the professional first step — we don't speak for any agency. What Unicaption itself does is designed for confidential work: it runs beside the call reading only your microphone or system audio, never joins or records anything, never stores audio, deletes transcripts at session end, and is end-to-end encrypted with nothing retained after the session.
Yes. Because Unicaption captures your microphone or system/tab audio rather than integrating with any specific platform, the same setup covers over-the-phone calls and video sessions alike — it sits beside Zoom, Teams, Google Meet, Webex, or your agency's own platform without anything joining the session. It runs in the browser, as a Chrome extension, and as native macOS and Windows desktop apps.
Unicaption ships with built-in medical, legal, and government terminology and supports 60+ languages with automatic detection and side-by-side translation. On top of that, a custom dictionary lets you add the specific drug names, clinic names, and regional terms your own queue produces — and it persists across sessions, so accuracy on your recurring vocabulary improves the longer you use it.
There are 30 free minutes every week, no credit card required, renewing weekly. Paid plans run from Instant at $9/month to Ultra at $149/month, with annual discounts. For a working interpreter, the sensible path is a few weeks of free minutes on real calls to see your true usage before subscribing.
30 free minutes every week, no credit card. Audio never stored, transcripts deleted at session end.
Start Free Trial →