VoiceFill
VoiceFill turns speech into finished documents in real time: it dictates structured medical reports for clinicians and auto-fills web forms during support calls. Built by Slovenian studio zerodays, it runs on web, iOS and Android in over fifty languages.
What is VoiceFill?
VoiceFill is a speech-to-text tool from the Slovenian software house zerodays d.o.o. that turns spoken dictation into a finished written document. The brand covers two surfaces. The original one, at voicefill.ai, is aimed at contact centres: it transcribes a live call, detects the fields of a web form and fills them in from what the caller says, then produces a summary once the call ends. The second, VoiceFill Medical, is the one the publisher actively maintains — the robots.txt and the sitemap of the main domain both point to it alone — and it addresses clinicians who dictate reports.
The medical product converts speech into structured text in real time, formats it as a report and lets the author edit the wording as it appears. Output can be exported to PDF or Word, using a practice's own template and preferred writing style. Punctuation and sentence structure are recognised automatically, so nothing has to be spelled out word by word. The publisher's stated starting point is that mainstream dictation engines handle Slovenian poorly, and closing that gap is the product's original purpose.
It runs as a web application and as published iOS and Android apps, both released by zerodays itself; sign-in can reuse an existing Google or Apple account. Transcription is advertised for more than fifty languages, of which thirty-two are named, while the interface exists in English and Slovenian. Dictaphones and external recorders are supported, and the vendor says it can connect to many existing reporting systems and EMRs, case by case.
Commercially, a permanent free tier allows ten reports a month; a Pro plan at 34.99 EUR per month lifts that ceiling and adds branding; an Enterprise tier, priced on request, adds API access, user management, an admin console, on-site onboarding and storage of audio recordings. The terms also provide a thirty-day free trial.
Two points deserve attention. The privacy policy commits, in plain words, to not using captured audio or transcripts to train machine-learning or generative models. Against that, the terms state the service is not tailored to HIPAA or FISMA, and no security certification is published.
What it does
- Dictate a medical report and get it back formatted in seconds
- Transcribe a call in real time without taking notes
- Fill a web form automatically from what is said in the conversation
- Generate a concise summary of a call or a transcript
- Export the result to PDF or Word using your own template
- Be prompted about questions not yet asked during a call
- Edit and correct any field before validating the document
When to use VoiceFill / When not to
A quick filter to help you decide if VoiceFill is the right fit.
When to use VoiceFill
- Doctors and clinicians who dictate findings and want a formatted report back in seconds
- Clinics and medical practices whose administrative staff retype dictated reports
- Slovenian-speaking healthcare teams poorly served by mainstream dictation engines
- Contact centre agents who must fill in a form while keeping the conversation going
- Medical software vendors looking to embed dictation through the Enterprise API
When not to use VoiceFill
- Organisations bound by HIPAA or FISMA, which the terms explicitly exclude from using the service
- Buyers who need a named hosting country or a signed DPA before processing patient data
- Anyone under 18, who the terms bar from registering
- Teams needing a certified supplier, as no ISO 27001, SOC 2 or comparable audit is published
- Contact centre buyers wanting a public price, since that version is quote-only
How to use VoiceFill
A typical end-to-end flow, from setup to results.
- Open the web app at app.medical.voicefill.ai, or install the iOS or Android app
- Sign in, reusing an existing Google or Apple account if you prefer
- Check you are online, as real-time transcription needs a connection
- Start dictating; speech is converted to text as you speak
- Let the system structure and format the report, without dictating punctuation
- Edit the text live, or correct any auto-filled field before validating
- Apply your own template and preferred writing style
- Export the finished document to PDF or Word
- On Enterprise, ask the vendor to connect your EMR, CRM or ERP, or to open API access
- For an organisation, request a tailored implementation plan and on-site onboarding
Pros & Cons
Pros
- The privacy policy rules out training models on captured audio and transcripts, in explicit terms
- A permanent free tier of ten reports a month, already including custom templates and PDF or Word export
- A single, readable entry price of 34.99 EUR per month, plus a thirty-day free trial in the terms
- A verifiable European publisher: the VAT number checks out and matches the address and the register
- Wide language coverage, with deliberate work on Slovenian that larger engines handle poorly
- Genuine iOS and Android apps, published by the company itself rather than a third party
- The FAQ answers directly on EMR integration, dictaphones and API access
Cons
- The terms exclude uses subject to HIPAA or FISMA, which sits awkwardly with a healthcare pitch
- The FAQ claims healthcare compliance while the terms say the opposite — the site contradicts itself
- No certification of any kind is published; end-to-end encryption is asserted but not evidenced
- No hosting country is disclosed and the hosting provider is named only by function
- Sub-processors are listed by category, so almost none can actually be identified
- No DPA is published or offered, which is awkward for a product handling patient data
- API documentation is not public and comes only on request
Pricing & Plans
A permanent free plan is available, limited to ten reports per month. The cheapest paid plan is Pro at 34.99 EUR per month. The currency is confirmed by the terms, which state that all payments are made in euros, and by the pricing data published on the site; a figure quoted in US dollars in one secondary file is not supported by either. The Enterprise plan is priced on request. A thirty-day free trial is provided for in the terms, after which the chosen subscription is charged. No public price is published for the contact centre version of the product.
- 10 reports per month
- your own unbranded template
- premium AI features
- PDF and Word export
- basic report editor
- email support answered in 48 to 72 hours
- unlimited reports
- customisable reports with your own branding
- premium AI features
- PDF and Word export
- basic report editor
- priority email support answered within 24 hours
- custom template
- dedicated phone support
- user rights management across the organisation
- on-site onboarding and training
- workflow integration with API access and CRM or ERP connections
- admin console with centralised billing
- storage and replay of audio recordings
Data, GDPR & hosting
A consolidated view of how VoiceFill handles your data.
GDPR overview
GDPR implementation is concrete and documented. The privacy policy, version 1.0 in force since 7 May 2026, names zerodays d.o.o. as controller, gives its address and contact address, and sets out purposes, legal bases, categories of data and retention periods in a table. The legal bases used are contract, law, legitimate interest and consent. Data subject rights are listed individually: access, rectification, restriction, objection, erasure, withdrawal of consent and portability, with erasure within thirty days of a withdrawal. The Slovenian supervisory authority, the Informacijski pooblaščenec, is named as the route for complaints. Transfers outside the EEA are acknowledged and covered by standard contractual clauses. One nuance: the site never states in so many words that it is GDPR compliant, so no compliance claim is recorded here. The Slovenian version of the policy prevails.
Who owns the data?
The publisher, zerodays d.o.o. of Ljubljana, splits its role in two. When it runs the service itself it acts as a processor: it handles the audio and transcripts on the customer's behalf, and forwards any request coming from a patient to the customer, who remains the controller. For its own purposes — billing and the subscription, newsletters, inbound enquiries, fraud prevention — it is the controller in its own right. Data is shared with a set of third parties described by function, some of which move it outside the EEA under standard contractual clauses. The controller can be reached at info@zerodays.dev, and rights or consent withdrawal at info@voicefill.ai.
Reuse rights
Customers keep the use of what they dictate: the reports and transcripts are produced for them, and can be edited before validation, then exported to PDF or Word and reused freely in their own practice or case files without asking the publisher for permission. On the Enterprise plan, audio recordings can be stored and replayed. The publisher's own reuse is expressly limited on the point that matters most: captured audio and transcripts are not used to train machine-learning or generative AI models. Beyond delivering the service, processing is confined to the purposes listed in the privacy policy, each with its own legal basis and retention period. Note that the terms reserve intellectual property in the software and the brand to the publisher, and restrict the licence to internal business use.
Data retention & training
Hosting summary
The site publishes no hosting location. The privacy policy identifies the hosting provider only by function, alongside the other third parties it relies on — a backend provider, a speech-to-text provider, subscription and payment management, crash reporting, bulk email, and the Google and Apple stores. None of them is named, and no country or region of processing is stated anywhere on the site. The policy does acknowledge that some of these third parties move personal data outside the EEA, and says this is covered by standard contractual clauses; that is a transfer safeguard, not a statement of where data sits. The IP address the domain resolves to is an anycast CDN node and says nothing about where processing happens. On the Enterprise plan, audio recordings can be stored and replayed, again without a stated location. A buyer with a data residency requirement — likely in a clinical setting — will have to obtain these commitments contractually, because the public documentation does not provide them.
Things to keep in mind
Risks and trade-offs to weigh before adopting VoiceFill.
- The site contradicts itself on healthcare compliance: the FAQ reassures, the terms exclude HIPAA and FISMA uses outright — read the terms before relying on it clinically
- A dictated report still carries the clinician's signature and liability; automatic formatting can read as authoritative while quietly dropping or mangling a detail that was said
- Real-time transcription of a consultation or a support call captures third parties who never consented — the patient or the caller — and that duty stays with the user
- Security claims such as end-to-end encryption and enterprise-grade protection are asserted without any certificate or audit to check them against
- No hosting country is published and sub-processors are named only by function, so data can leave the EEA without the user knowing which supplier holds it
- Leaning on automatic capture erodes the habit of active listening and note-taking, and the loss only shows when the tool is unavailable or mishears
- The free tier of ten reports a month is easy to outgrow mid-workflow, which turns an evaluation into a purchase decision sooner than expected
Setup & Integrations
Technical difficulty
Low for an individual user. The product opens in a browser at the web app, or through the iOS and Android apps, and sign-in can reuse an existing Google or Apple account; dictation can start immediately, with no installation described. An internet connection is required, since transcription runs in real time. Effort rises only for organisations: connecting an EMR, CRM or ERP, or using the API, goes through the vendor's team on the Enterprise plan, with on-site onboarding and a tailored implementation plan. The publisher describes roll-out as fast, with the duration depending on existing infrastructure and the integrations required.
Deployment
Apps stores
Supported languages
Behind VoiceFill
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does VoiceFill actually do?
How much does it cost?
Are my recordings used to train AI models?
Is it suitable for a HIPAA-regulated organisation?
Which platforms and languages are supported?
Can it connect to my existing reporting system or EMR?
Is there an API?
Where is my data hosted?
How long is data kept?
Who is behind VoiceFill and is there a minimum age?
Should you pick VoiceFill?
VoiceFill is a focused, unpretentious tool from a Slovenian micro-company of around eight people, and it is best judged on that scale rather than against the large dictation platforms. What it does, it does clearly: turning speech into a formatted report or a filled-in form, in real time, on the web and on both mobile platforms, with a genuine effort on a language that bigger engines serve badly. Two things stand out commercially — a permanent free tier that is usable rather than token, and a single readable price of 34.99 EUR per month above it.
The strongest signal of seriousness is the explicit commitment not to train models on captured audio or transcripts. That is a real differentiator, and one stated plainly rather than buried.
The reservations are documentary rather than functional. A product sold to clinicians publishes no certification, names no hosting country, identifies its sub-processors only by function and offers no data processing agreement. More awkwardly, the site contradicts itself: the FAQ speaks of healthcare compliance while the terms exclude any use subject to HIPAA or FISMA. The public-facing landing page for the contact centre version, meanwhile, is an unedited template that no longer reflects where the product is going.
For an individual clinician or a small practice — especially a Slovenian-speaking one — the free tier makes this cheap to evaluate on real work, and the answer will come quickly. For a hospital or any organisation with a procurement process, the gap to close is contractual: a signed DPA, named sub-processors and a hosting commitment would all have to be obtained directly from the vendor, because none of them is on the website today.
- Choosing a selection results in a full page refresh.
- Opens in a new window.