Priberam Care
Clinical AI suite from Portuguese NLP publisher Priberam, combining three modules that turn consultations into SOAP reports, suggest explainable ICD-10 codes and draft discharge notes, with every model runnable inside the hospital itself.
What is Priberam Care?
Priberam Care is a clinical documentation platform built by Priberam, a Portuguese natural language processing company based in Lisbon. It groups three distinct modules under one brand, each aimed at a different part of the paperwork that surrounds a medical episode.
Assistant listens to the consultation itself and writes the report. The doctor starts a recording, talks with the patient as usual, then reviews the generated SOAP report and copies it into the record. While the conversation is running, the module gives live feedback on what it is capturing, and it applies terminology glossaries, both Priberam's own term lists and any list the practice adds. Only the SOAP format is supported today; further templates are announced.
Codes serves medical coders rather than clinicians. It gathers the documents belonging to a medical episode, proposes ICD-10-CM/PCS codes, and explains each one by attributing the passage of text that justifies it. The coder selects an episode, reviews the documents, checks the suggested codes and their explanations, then saves and submits. The interface was designed alongside working coders, and the same workflow supports auditing codes that already exist.
Notes generates discharge notes. It can build a clinical history from a patient's earlier records, summarise inpatient stays and outpatient surgeries, and cope with patients whose files are long. It is also positioned for emergency admission, where it retrieves previous diagnoses and usual medication automatically.
The argument running through all three is time. Priberam states that physicians spend roughly 45% of their working time on the electronic health record and report symptoms of burnout. The company frames the product around four Responsible AI pillars, sustainability, privacy, explainability and fairness, and makes an unusual technical commitment: every model can run inside the customer's own clinic, and Priberam hosts only where the organisation lacks the resources to do so. It explicitly claims no dependence on OpenAI or other generative AI providers.
Hospital de Sao Joao and Hospital da Luz are shown as organisations already using the platform. There is no public pricing, no free trial, no mobile application and no documented public API: the only way in is the contact form.
What it does
- Transcribe a doctor-patient consultation and write the SOAP report from it
- Suggest ICD-10-CM/PCS codes for a medical episode and justify each one
- Generate discharge notes for inpatient stays and outpatient surgeries
- Assemble the documents of an episode for coding or audit
- Build a clinical history from a patient's past medical records
- Retrieve previous diagnoses and usual medication when a patient is admitted
- Apply custom terminology glossaries to transcription and reporting
When to use Priberam Care / When not to
A quick filter to help you decide if Priberam Care is the right fit.
When to use Priberam Care
- Physicians who dictate or write up consultations, inpatient stays and outpatient surgeries
- Medical coders assigning or auditing ICD-10-CM/PCS codes across hospital episodes
- Hospitals and clinics that require models and patient data to stay on their own infrastructure
- Emergency departments needing previous diagnoses and usual medication pulled from past records at admission
- Portuguese-speaking healthcare organisations, given the publisher's specialism in the language
When not to use Priberam Care
- Buyers who need a published price, since nothing is quoted and every route runs through a sales conversation
- Individual practitioners wanting to sign up and start the same day, as there is no self-service onboarding
- Teams expecting a mobile app, because neither an iOS nor an Android application exists
- Developers looking for a documented public API for the product itself
- Organisations that require published certifications, a product privacy policy or stated retention periods before procurement
How to use Priberam Care
A typical end-to-end flow, from setup to results.
- Contact Priberam through the form, as there is no self-service sign-up
- Agree with the publisher whether the models run in your clinic or on Priberam's infrastructure
- In Assistant, start the recording at the beginning of the consultation
- Hold the consultation normally while the module signals what it is capturing
- Review the generated SOAP report and copy it into the patient record
- In Codes, select the medical episode you need to code or audit
- Review the documents gathered for that episode
- Check the suggested codes and the text passages given as explanation
- Save and submit the codes
- In Notes, pick the episode and a template, then review and copy the discharge note
Pros & Cons
Pros
- Models can run inside the hospital, a rare commitment for clinical AI handling patient data
- Explicit independence from OpenAI and other generative AI providers
- Explainability built into the product, with text spans attributed to each predicted code
- Three complementary clinical workflows covered under a single vendor
- Coding module designed together with practising medical coders
- Established publisher with a verifiable legal identity, address and VAT number in Lisbon
- No invented credentials: the site claims no certification it cannot support
Cons
- No public pricing at all, so the cost cannot be estimated without contacting sales
- No privacy policy covering the product or the handling of patient data
- The legal texts are unedited British boilerplate on a Portuguese publisher
- Two unedited theme template pages remain live, one showing fictitious packages at $0
- No stated retention periods, no subprocessor list and no data processing agreement
- No named EHR integrations and no declared processing languages
- No mobile application and no documented public API
Pricing & Plans
No price is published for Priberam Care. There is no pricing page, no free trial and no free plan; the only calls to action are Contact us and Talk to us, both leading to a contact form. The product is therefore sold on quotation and no entry price can be stated. A page reachable only through the sitemap shows three packages at $0 per month, but it is an unedited theme template carrying placeholder text, a fictitious patient review and a foreign placeholder telephone number, and it describes no real offer.
Data, GDPR & hosting
A consolidated view of how Priberam Care handles your data.
GDPR overview
The publisher's privacy policy places itself explicitly under Regulation (EU) 2016/679 and states that Priberam acts as both Controller and Processor. It offers objection to marketing processing, erasure, rectification, and a full copy of the data held as a PDF within thirty days. Because Priberam is established in Portugal, no Article 27 representative is required and none is named; no data protection officer and no dedicated privacy address are published either. The site never claims GDPR compliance in so many words, so no compliance flag is asserted here. One caveat matters: the text is an unedited British boilerplate, citing section S11596 of the Companies Act 2006 and referring to women in SW1, which sits oddly on a Portuguese publisher and suggests it was never reviewed against the product.
Who owns the data?
Priberam Care publishes no privacy policy of its own. The only document available is the privacy policy of the publisher's corporate site, and it governs that website rather than the clinical data the product handles. In it, Priberam declares itself both Controller and Processor under the GDPR. It grants the right to have all information relating to a person removed from its systems, and the right to receive a full copy of the data held, supplied as a PDF within thirty days. Nothing in the collected pages states who owns the consultation recordings, transcripts, generated reports or assigned codes, nor what Priberam may do with them.
Reuse rights
Nothing in the collected material grants or limits a right for the customer to reuse what the product produces. The publisher's privacy policy describes website uses only: information given through forms, correspondence, surveys, transaction details and visit logs, plus cookies for audience measurement and preferences. It permits data to be transferred and stored outside the European Economic Area, and allows selected third parties to be given data for marketing, with an opt-out by tick box. No page states whether recordings, transcripts, discharge notes or assigned codes may be exported, redistributed or reused by the customer. Priberam does say its models can run inside the clinic without dependence on OpenAI or other generative AI providers, but that is a statement about hosting, not a licence over the output.
Data retention & training
Hosting summary
Priberam makes one clear hosting claim: all the models used can be hosted in your clinic, and if the organisation does not have the resources for that, Priberam will host them itself. It also states that there is no dependence on OpenAI or other generative AI providers. Beyond this, nothing is specified. No hosting country and no region is named anywhere for the product, so no jurisdiction can be recorded. The publisher's website privacy policy does mention that data it collects may be transferred to and stored outside the European Economic Area and processed by staff operating outside it, but that clause concerns the corporate website and is not a statement about where the clinical platform runs. The marketing site itself is served from Lisbon, which describes the brochure rather than the product. Prospective customers should settle hosting location, jurisdiction and subprocessors contractually, since the public pages do not answer these questions.
Things to keep in mind
Risks and trade-offs to weigh before adopting Priberam Care.
- Automation bias: a fluent SOAP report or a confidently suggested code invites rubber-stamping, and the clinician remains responsible for what enters the record
- The site publishes no retention period, no subprocessor list and no product privacy policy, so the fate of consultation recordings and transcripts is undocumented
- Ambient recording of consultations raises patient consent questions that the site never addresses
- Deskilling risk: coders and juniors who stop composing notes and codes themselves may lose fluency they need when the tool is unavailable or wrong
- The legal pages are unedited foreign boilerplate, so they cannot be relied on to describe the actual contractual position
- Two live template pages show fictitious packages, prices and reviews, which can mislead a reader into thinking terms are published
- Nothing states whether customer or patient data is used to train models, so the question must be settled contractually rather than assumed
Setup & Integrations
Technical difficulty
Getting started is an organisational project rather than a technical one for the end user. There is no self-service sign-up: everything begins with a sales conversation, and the deployment model must be agreed, with models running either inside the clinic or on Priberam's infrastructure. Integration with the patient record is implied, since the modules select episodes and read past documents, but no connector is named and no technical documentation is public. Day-to-day use is straightforward once installed: each module follows four steps and ends with copying the result.
Deployment
Behind Priberam Care
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does Priberam Care actually do?
Is Codes the whole product?
Where do the models run?
Does it depend on OpenAI?
How much does it cost?
Is there a free trial?
Is there a mobile app or a public API?
Which report formats does Assistant produce?
Who is behind it?
Which organisations already use it?
Should you pick Priberam Care?
Priberam Care is a credible clinical documentation platform rather than a speculative one. It comes from an established Portuguese publisher whose legal identity, Lisbon address and VAT number all check out, it names two real hospital groups as users, and it takes a position most competitors avoid: the models can run on the hospital's own infrastructure, with no dependence on OpenAI or other generative AI providers. For patient data, that is a substantive argument rather than a marketing one, and the explainability built into the coding module, where each suggested code is tied back to the text that justifies it, is the kind of feature coders can actually audit.
The reservations are about the shop window, not the technology. There is no published price, so the budget conversation can only happen with a salesperson. More seriously for a product handling clinical records, the site publishes no privacy policy of its own, no retention period, no subprocessor list and no data processing agreement; the only policy available covers the publisher's corporate website and is an unedited British boilerplate that cites UK company law. Two unedited theme template pages are still live, one of them advertising fictitious care packages at $0 a month, and a placeholder Gmail address appears in the markup of every page. None of this is evidence of a bad product, but it does mean the public site answers far fewer questions than a hospital procurement process will ask.
Treat the website as an introduction and put the real questions to the vendor: contractual terms, where data is held, how long it is kept, which EHR systems are supported, and what independent assurance exists. The technical proposition is worth that conversation.
- Choosing a selection results in a full page refresh.
- Opens in a new window.