voize
voize is an AI voice assistant for care documentation. Nurses speak at the point of care and the app structures what they say into entries that sync into the facility's existing EHR, cutting documentation time by a claimed 27%.
What is voize?
voize is a voice documentation application built for long-term care, skilled nursing, hospital wards and disability support services. The premise is narrow and deliberate: instead of walking back to a workstation at the end of a shift, a caregiver speaks into a smartphone during or immediately after the care episode, and the AI turns those words into structured entries filed in the record system the facility already uses. voize does not replace the EHR — it writes into it, in both directions, and adds a spoken layer on top.
What separates it from a generic dictation tool is the model underneath. The vendor states that it built its own domain-specific language model trained on real care processes rather than relying on a general-purpose LLM such as ChatGPT, and that this model understands professional terminology, copes with dialects and helps caregivers who are not native speakers. It structures speech into the documentation formats practitioners actually use, recognises resident names and assigns each entry to the right file, and corrects grammar and punctuation as it goes. Coverage is broad: vital signs, sign-off of planned interventions, ad-hoc measures and free-text care reports, structured assessments for pain, mobilisation, wounds and falls, handover notes and complex scales such as Barthel and Braden.
The application works offline, which the vendor presents as unusual for voice AI and decisive in buildings where wireless coverage is patchy; entries sync in the background once a connection returns. A caregiver reviews and approves every draft on the device before it is filed, so a human stays in the loop.
The claimed results are specific: 27% less time spent on documentation, measured in a study run with the Charité, alongside 15% lower staff turnover and reduced overtime. Adoption figures cited across the site range from 1,800 to more than 2,000 facilities and around 180,000 caregivers, with more than 200 million entries documented. The product is sold in Germany, Austria, France and the United States, and carries the compliance apparatus each market expects: GDPR and HDS in Europe, HIPAA and a Business Associate Agreement in the United States, plus the German church data-protection regimes. Newer capabilities include an AI agent for care planning, digital shift handover and phrasing assistance.
What it does
- Dictate a care observation on a smartphone, at the resident's bedside, during the visit
- Turn spoken notes into structured entries mapped to the right fields of the existing record system
- Capture vital signs, planned interventions, ad-hoc measures and free-text care reports by voice
- Complete structured assessments: pain, mobilisation, wound documentation, fall protocols, Barthel and Braden
- Review, correct and approve every entry before anything is filed
- Keep documenting with no network and let entries sync in the background once connectivity returns
- Push confirmed entries into the EHR in real time, visible across shifts, profiles and devices
When to use voize / When not to
A quick filter to help you decide if voize is the right fit.
When to use voize
- Long-term care and skilled nursing facilities where caregivers lose hours of every shift to charting
- Hospital wards that want spoken notes to reach the existing record system without a second data entry
- Home care and community services documenting on the move, in buildings where the network is unreliable
- Care providers preparing for PATHOS, PDPM or CMS-style reviews who need complete, timely records
- Operators trying to cut unplanned overtime and staff turnover among nursing teams
When not to use voize
- Organisations shopping for an electronic health record: voize adds a voice layer, it does not replace the EHR
- Individual clinicians and freelancers: licensing is per facility, based on the number of people cared for, never per personal account
- Buyers who need a published price before speaking to a salesperson
- Teams expecting to sign up and get going alone: there is no free plan, no announced trial, only a booked demo
- Facilities without a managed handset fleet, since phones are owned by the provider and enrolled in mobile device management
How to use voize
A typical end-to-end flow, from setup to results.
- Agree a scope with the vendor after a demo, since there is no self-service sign-up
- Have the integration set up against the existing record system through its API, with no change of platform
- Provide the handsets: any iPhone or Android from the last five years, owned by the facility
- Work with the IT department to enrol those phones in mobile device management, so they stay restricted to work use
- Log in on the mobile device to see resident profiles, care plans, medications and tasks at a glance
- Speak naturally during or right after the care episode, instead of memorising and typing later
- Let the AI structure what was said, recognise the resident's name and route it to the right fields
- Review the draft on the phone, edit anything that was misheard, then approve it
- Rely on offline mode when the network drops, and let entries sync automatically afterwards
- Watch approved entries appear immediately across shifts, profiles and devices for the whole team
Pros & Cons
Pros
- A measured time saving rather than a marketing figure: 27% less documentation time in a study run with the Charité
- Offline operation that actually holds up, which matters in buildings with unreliable wireless coverage
- Integrates into the record system already in place instead of demanding a migration
- In the standard US deployment, speech recognition and documentation are processed on the caregiver's device, so protected health information stays out of the vendor's cloud
- A human stays in the loop: nothing is filed until the caregiver has read and approved it
- Compliance is addressed on three fronts at once, covering GDPR and HDS, HIPAA with a signed Business Associate Agreement, and the German church data-protection statutes
- A large, named customer base including Caritas, Diakonie, Evangelische Heimstiftung, Newcare and Bridgeway Senior Healthcare
Cons
- No published price anywhere: the pricing page offers only a request form, with no range and no indicative figure
- No free plan and no announced free trial; the only entry point is a sales demo
- No terms and conditions published on the entire site, so the contract cannot be read before making contact
- An API and an SDK are claimed, in the FAQ and in the vulnerability programme, yet no public documentation for either exists
- The product privacy policy is the US one, declares itself subject exclusively to United States law and states that servers are located in the United States, which sits awkwardly beside the GDPR and HDS messaging on the European pages
- The German and French privacy notices cover webinars only, and the French page serves the untranslated German text
- No subprocessor list is published, and no mechanism is documented for excluding your data from model training
Pricing & Plans
voize publishes no prices. There is no permanent free plan and no announced free trial; the only self-service step on the site is booking a free demonstration. Licensing is quoted rather than listed, and is based on the number of people cared for rather than on user accounts, so every caregiver in a facility may use the application without additional cost, with no penalty for part-time staff, floaters or turnover. Every pricing route on the site leads to a request form. The German pages set out four recognised ways for a provider to absorb the investment within existing budgets: as a staffing equivalent recognised in remuneration negotiations under §85 SGB XI, as a running operating cost, as capital expenditure, or through savings on overtime, agency staffing and turnover. A quotation must therefore be obtained directly from the vendor.
Data, GDPR & hosting
A consolidated view of how voize handles your data.
GDPR overview
GDPR compliance is claimed explicitly and in several places. The French pages state that voize complies with the GDPR and that core services run on HDS-certified infrastructure meeting French health-data requirements; the funding announcement repeats that all data is processed under the GDPR. A data processing agreement is concluded at contract signature, aligned with the GDPR and with the German church data-protection statutes EKD and KDG. voize GmbH is named as controller, an external data protection officer is appointed (Althammer & Kill, Hanover), the competent supervisory authority is the Berlin data protection commissioner, and rights of access, rectification, erasure, restriction and portability are exercised at info@voize.de. One gap matters: the German and French privacy notices cover webinars only, and the product privacy policy is the US one, which declares itself subject exclusively to United States law.
Who owns the data?
The site never spells out in so many words who owns the dictated content, but the US privacy policy draws the boundaries. voize positions itself as a HIPAA Business Associate: protected health information is used and disclosed only as the Business Associate Agreement, the underlying service contract and HIPAA permit, and never for voize's own commercial purposes. The vendor undertakes not to trade, rent, share or sell personal information, and restricts access to the employees, service providers and subcontractors who need it. One reservation deserves attention: voize keeps the right to use anonymised and de-identified data, including de-identified health data, and to pass such aggregate data to third-party collaborators and partners.
Reuse rights
Nothing published on the site limits what a customer may do with its own records, which fits the design: confirmed entries are written into the facility's own EHR, where they remain under the provider's control and its usual clinical governance. Caregivers review, edit and approve every entry before it is filed, and can still amend it later in the host system. The caveat is documentary rather than technical. voize publishes no terms and conditions anywhere on the site, so there is no licence text, no export clause and no reuse restriction available to read before contracting. Any organisation that needs written confirmation of its reuse rights will have to obtain it from the vendor during negotiation.
Data retention & training
Hosting summary
The site tells two different stories depending on which version you read, and neither is complete. The US privacy policy, which is the only privacy notice covering the product, states plainly that voize servers are located in the United States and that the policy is subject exclusively to United States law. The US FAQ adds that AWS is the underlying cloud provider, that a Business Associate Agreement with AWS has been executed, and that where US backend services are needed they run in regions chosen to align with HIPAA expectations. It also stresses that the standard US deployment keeps protected health information out of the vendor's cloud altogether, with speech recognition and documentation processed on the caregiver's device and structured notes sent straight into the customer's record system. The French pages take a different line, saying core services run on HDS-certified infrastructure under French health-data requirements, with part of the processing done on the mobile device. HDS is a certification, not a location: no European hosting country is named anywhere. A European buyer should get the answer in writing.
Things to keep in mind
Risks and trade-offs to weigh before adopting voize.
- Rubber-stamping: the human-in-the-loop safeguard only works if caregivers genuinely read the draft. Under time pressure, approval becomes a reflex and an AI misreading is filed as clinical record
- Skill erosion: writing a care observation is itself an act of clinical reasoning. Delegating the phrasing for years may quietly weaken a team's ability to structure and articulate what it observes
- Documentation quality feeds reimbursement here, through PATHOS, case mix and PDPM. Anything that makes records fuller also makes them more lucrative, which is a standing invitation to drift toward over-documenting
- It is not obvious which legal entity a customer contracts with: the imprint names voize GmbH in Berlin, while the product privacy policy is signed by a US entity and points to a New York address and a separate domain
- The hosting story differs by language version. The US policy states servers are in the United States; the European pages lead with HDS certification and on-device processing, and name no country at all
- No opt-out from model training is documented, while the vendor states its models keep learning from real usage and reserves the right to share de-identified health data with third-party collaborators
- Dictation happens at the bedside, so speech is captured in a space shared with residents and colleagues. How consent and bystander privacy are handled locally is a question for the provider, not the software
Setup & Integrations
Technical difficulty
Moderate, and mostly organisational rather than technical. The vendor handles the software side: an API integration into the record system already in place, with no change of platform and two-way sync. The work falls on the facility, which must supply and own the handsets, then involve its IT department to enrol them in mobile device management so the phones stay restricted to work use. No phone lines are needed and wireless is enough, since the application also runs offline. Customers commonly start with a pilot. No public technical documentation exists, so scoping must be done with the vendor.
Deployment
Integrations
Supported languages
Behind voize
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is voize?
Does voize replace our record system?
How is it integrated?
Does it work without a network connection?
What devices are needed?
What happens if the AI mishears something?
Which documentation tasks does it cover?
Is it GDPR compliant, and what about HIPAA?
What does it cost?
What time saving is actually demonstrated?
Should you pick voize?
voize is an unusually disciplined product: it does one thing, which is to let care staff document by speaking, and it resists the temptation to become a record system of its own. That restraint is its main strength. Integrating into the EHR a facility already runs removes the migration argument that kills most care-sector software projects, and the offline capability answers the objection that actually stops voice tools working in older buildings.
The evidence offered is better than the sector average. A 27% reduction in documentation time measured with the Charité, a named customer base spanning Caritas, Diakonie and Evangelische Heimstiftung, a USD 50 million Series A led by Balderton Capital, and a proprietary model trained on real care processes rather than a general-purpose LLM all point to a company past the prototype stage.
The reservations are about transparency rather than capability. Nothing on pricing is public, and no terms and conditions exist anywhere on the site, so a buyer cannot read the contract before making contact. An API and an SDK are claimed without any documentation to back them. Most significantly, the product privacy policy is a United States document, signed by a US entity, declaring itself subject exclusively to US law and stating that servers are located in the United States, while the European pages lead with GDPR and HDS. The German and French privacy notices only cover webinars. A European buyer should ask, in writing, which legal entity it is contracting with and where care data physically resides, before anything else.
For a care provider drowning in charting, voize is worth the demo. Go into it with those questions prepared.
- Choosing a selection results in a full page refresh.
- Opens in a new window.