Amdia, the smart care assistant
More room for good care, from intake to follow-up.
Amdia is a smart care assistant that thinks along with care provider and patient. At home, your patient shares what is going on and is supported along the entire care pathway. You get work off your hands and keep more joy in your profession.
- 125+care pathways with a validated knowledge base
- 30+care organisations in our network
- 15years active in hospitals and clinics
Built on quality. Strengthened by AI.
For 15 years, we have been working with care providers on better communication between patients and care providers, based on validated knowledge bases. With well-applied AI, we now bring even better experiences to more patients, at more moments.
Choose your way in
For care providers
A smart assistant that thinks along: prepared patients and more room for good care.
Read on
For patients
Read how Amdia works and what happens to your answers.
How it works for you
For health insurers
Decision aids for your members, and appropriate care across your care network.
How it works for youOne smart assistant, three parts
One smart assistant, along the entire care pathway.
For all contact between care provider and patient. Starting with intake, explanation and decision aids, and follow-up.
- Before the appointment
Intake
At home, your patient shares their symptoms and context; you receive an overview beforehand.
How it works - Before and after the appointment
Explanation and decision aid
Validated explanations; when there is a choice, the decision aid sets out the options.
How it works - During follow-up
Follow-up
At set moments, your patient shares how things are going, before the next appointment.
How it works
A validated knowledge base for every care pathway
Validated by medical specialists and kept up to date. All three parts run on it.
- Guidelines
- Literature
- Validated intake question structure
- Local process and patient information
- Validated decision aids
Why Amdia
This is how we earn your trust.
Not a chatbot that guesses. A validated knowledge base.
Every answer comes from a knowledge base validated by medical specialists, with source, version and date. The intake starts from a fixed set of reviewed questions. Every new version is tested before it goes live.
Not a separate tool. One assistant along the entire care pathway.
Years of experience in Dutch hospitals and clinics taught us that separate tools cost care providers extra work. That is why Amdia guides the patient from intake to follow-up.
Not a promise. Evidence.
With our decision aids, patients more often chose what suits them, with 13 to 15 percentage points fewer operations for gallstones and inguinal hernia, and therefore more appropriate care. Now we are studying the smart assistant in five care pathways.
Intake and follow-up
A prepared care provider, at every appointment.
At home, the patient shares their symptoms and context: before the first appointment with the intake, and then at set moments with the follow-up. So you know beforehand why the patient is coming and how things are going, and the conversation starts with what matters.
An invitation from the patient's own hospital
At a chosen moment in the care pathway, through a link from the hospital EHR or the GP information system. No password needed.
The patient shares, at their own pace
At home, on the go or in the waiting room. One short question at a time, saved automatically. After the appointment, the patient shares at set moments how things are going.
The care provider reads it before the appointment
An overview of what the patient shared about symptoms, history, medication and expectations, checked by the patient. For a follow-up appointment: how things went since last time.
Intake · CardiologyEastfield Hospital
OverviewTopic 2 of 6 · Your symptoms Saved automatically
Intake assistant
Hello! I'll help you describe your symptoms clearly for your doctor. I ask one short question at a time.
What I do and don't do
- Record your symptoms step by step
- List your expectations
- Make a diagnosis or give medical advice
What is the reason for your appointment? Feel free to use your own words.
I've had palpitations for a few weeks, mostly in the evening on the sofa.
SYMPTOMPalpitations, for a few weeks, mostly in the evening on the sofa
Is that right, or would you like to add something?AI can make mistakes and is not a substitute for a healthcare professional.
What the care provider receives
- Reason for visit
- Palpitations, for a few weeks
- When
- Mostly in the evening at rest, a few minutes
- Medication
- None
- Allergies
- None known
- Wants to discuss
- Whether it is harmful, and whether exercise is OK
Summary in medical languageIn development
Reason for referral
Palpitations for several weeks, mainly in the evening on the sofa. Episodes of a few minutes, self-limiting.
History
No chest pain, no syncope, no exertional dyspnoea. Fatigue after an episode. Coffee 4–5 cups/day, alcohol 1 unit/day, non-smoker.
PMH / FHx
PMH unremarkable. FHx: father AF, diagnosed at age 65.
Medication / allergies
No medication. No known allergies.
Patient’s questions
Is it dangerous? Can she keep running?
Full summary: 6 of 6 sections
The summary in medical language for the care provider is in development and not yet available.
Example screens with a fictional hospital.
What it gives you as a care provider
Before the appointment, you know why your patient is coming and how things are going. That helps you:
-
Less administrative burden
In hospitals, 36% of working time goes to administration. What your patient shared at home is ready before the appointment.
-
More efficient appointments
In a randomised trial, the appointment was 29% shorter when a chatbot took the history beforehand. In Germany, too, history taking was faster. You know beforehand why your patient is coming, and keep more room in the appointment for what matters.
-
Less capacity needed
When check-ups are matched to how the patient is doing, fewer visits are needed at equal quality: 34% fewer in a Dutch trial. With the follow-up, you see who you really need to see.
Fits the existing workflow
Explanations and decision aids per care pathway
Reading up with their own care provider, not with Doctor Google.
Before or after the appointment, the care provider sends the patient a smart assistant about their condition. The patient reads up at their own pace: what the condition is, how it usually progresses and which treatments exist.
If a choice is coming up, the decision aid sets out the options, the pros and cons and what matters to the patient. All information is validated by medical specialists. The diagnosis is always made by the care provider, not the assistant.
Explanation · Knee osteoarthritisEastfield Hospital · from Dr Baker
My questions (3)Knee osteoarthritis assistant
Welcome. Dr Baker has prepared this explanation for you, ahead of your next appointment. What would you like to know?
Which treatments are there, apart from surgery?
ADDED TO YOUR LISTWhich treatment suits me?
Based on information validated by the orthopaedics department
AI can make mistakes and is not a substitute for a healthcare professional.
- Which treatment suits me?
- How long before physiotherapy helps?
- Can I keep walking and cycling?
I also want to mention
- The pain has got worse at night.
Example screens with a fictional hospital.
What it gives you as a care provider
A patient who understands their care makes your work lighter and care better.
-
Better care outcomes
Three in ten Dutch people struggle with health information, which is associated with more admissions and more mistakes with medicines. With explanations they can read back, patients know more.
-
Demonstrably appropriate care
With a decision aid, patients more often choose what fits what matters to them. In a Dutch study, their preference after the decision aid matched the eventual treatment in 86% of cases. What your patient fills in on the decision aid comes back to you.
-
A better conversation, less one-way talk
Your patient has already read the basics. In a randomised trial, the conversation was rated better with AI explanations beforehand, especially for shared decision-making, and was 21% shorter.
-
Fewer questions in between
Patients forget 40 to 80% of what you tell them. At home, they read it back and ask the smart assistant their questions first. In one study, a chatbot answered almost a third of the questions itself.
Reliable explanations, in line with your care.
Safe and responsible
The care provider decides, AI prepares.
The smart assistant does not diagnose and gives no treatment advice. Diagnosis and the choice of treatment rest with the care provider, together with the patient. The smart assistant says so upfront, too.
More about safety and privacyWhat the smart assistant does
- Record your patient’s symptoms step by step
- Record medication, medical history and allergies
- List your patient’s expectations and questions
What the smart assistant doesn't do
- Make a diagnosis
- Decide which treatment your patient gets
- Give medical advice
Developed with clinical practice
Not a solution from the drawing board: it grew out of clinical practice and was developed further together with care providers. The medical content is validated by medical specialists.
By invitation from their own care provider
Patients only use Amdia when invited by their own hospital or care provider, as part of their own care pathway.
Privacy and security from day one
Stored encrypted in the Netherlands, processed in the EU. Privacy, security and regulation are built in from the start. That saves long discussions with IT and your data protection officer.
One supplier, one platform
No separate tools or extra contracts. Intake, explanations and decision aids, and follow-up on one platform: less procurement, management and coordination with IT and security.
What we build on
Built on years of working together on better care.
Medical expertise, a network of healthcare professionals and high-quality technology: that is what Amdia builds on.
active in hospitals and clinics, improving patient information
healthcare professionals in our network
integrations available for hospitals
care pathways with a validated knowledge base
validated data points per care pathway, backed by literature
care organisations in our network
That actually makes the consultation more personal. Warm care, in other words.
Because patients are better prepared, we can talk better about the most suitable treatment. I really notice that I’m asked less and less: ‘What do you think I should do, doctor?’
During the consultation we get to the point faster. A prepared patient is the best person to talk with.
More experiences from practice
You can have a real conversation with the patient in the time that would otherwise go into giving all the information.
Programme manager, innovation and strategy, regional hospital Now that patients can look into the options beforehand, there is more time in the conversation and it’s easier to make a choice during it. That avoids follow-up appointments or postponed decisions.
Gynaecologist Patients need to get the right answers to their questions. And they want those answers when it suits them, simply at home on the sofa.
Gynaecologist Specialists benefit from the smart modules too: more job satisfaction because patients know more.
Programme manager, innovation and strategy, regional hospital This strengthens the relationship between doctor and patient and helps to set out all the considerations clearly.
Surgeon and principal investigator, university medical centre I actually see it as something positive. If patients have looked up and read information at home, they already have knowledge. That helps us think further together and deepen the conversation.
Gynaecologist We also notice that patients come to the clinic better prepared and start a treatment with more conviction.
Surgical medical assistants, regional hospital Everything is now in one place, linked to the patient in the EHR. It saves time, it's less error-prone and you no longer need to scan or print anything.
Surgical medical assistants, regional hospital In the past, as a doctor you said which treatment was best. Now we also help patients weigh things up for themselves beforehand. That way, patients feel better about the treatment they choose.
Surgeon and principal investigator, university medical centre Because with the huge staff shortage, we simply can’t do it on our own in healthcare any more.
Gynaecologist Involving patients in their own care pathway raises the quality of care and gives patients more control.
Internist-endocrinologist If you know what your options are, you feel less at the mercy of others. And if you know why you are choosing something, it leads to more trust in and commitment to the treatment.
Gynaecologist When you are told you have breast cancer, it is important to understand your situation. A good decision aid helps you understand your illness and the differences between treatments.
Project lead at a patient association
Read more experiences (13)
Impact on the care organisation
Free up capacity for higher-quality care.
These are five places where Amdia can take work away.
- Less administration per consultation
- More efficient consultations, because patients arrive better prepared
- Fewer phone and online questions to the clinic
- Fewer follow-up contacts for explanations
- Patients who already have their story and questions in order
You decide what to do with the time that frees up
Help more patients with the same team.
More time and attention for each patient.
Start small, expand easily. Begin with one care pathway and expand to other care pathways and specialties, on the same platform.
More room for good care starts with you.
With a care provider who is relieved of work, and a patient who gets a grip on their health. Do you believe in that too? Let’s make it happen together.
Source 1
Latenstein, C.S.S. et al. (2019). Reduced Elective Operation Rates and High Patient Satisfaction After the Implementation of Decision Aids in Patients with Gallstones or an Inguinal Hernia. World Journal of Surgery, 43(9), 2149–2156. Before-and-after study in one Dutch hospital: 15.3 percentage points fewer operations for gallstones and 13.2 for inguinal hernia.
Open the source (opens in a new tab)Source 2
Statistics Netherlands (CBS) (2025). Almost a third of working time in care goes to administration. AZW employee survey, Q2 2025. In hospitals and other medical specialist care: 36% of working time (record keeping 22.9%, registration 12.6%). In Dutch.
Open the source (opens in a new tab)Source 3
Tao, X. et al. (2026). An LLM chatbot to facilitate primary-to-specialist care transitions: a randomized controlled trial. Nature Medicine. 2,069 patients at two hospitals in China: consultation 3.1 vs 4.4 minutes (28.7% shorter) when a chatbot took the history beforehand.
Open the source (opens in a new tab)Source 4
Schneider, S. et al. (2023). Successful usage of a chatbot to standardize and automate history taking in Hymenoptera venom allergy. Allergy. Randomised pilot in Germany.
Open the source (opens in a new tab)Source 5
de Jong, M.J. et al. (2017). Telemedicine for management of inflammatory bowel disease (myIBDcoach): a pragmatic, multicentre, randomised controlled trial. The Lancet, 390, 959–968. 909 patients in Dutch hospitals: 1.55 vs 2.34 outpatient visits per year (34% fewer), with equal quality of care.
Open the source (opens in a new tab)Source 6
Knottnerus, B., Heijmans, M. and Rademakers, J. (2021). Inclusive COVID-19 prevention. Nivel (Netherlands Institute for Health Services Research). Three in ten Dutch people (29%) have difficulty finding, understanding and applying health information. In Dutch.
Open the source (opens in a new tab)Source 7
Berkman, N.D. et al. (2011). Low health literacy and health outcomes: an updated systematic review. Annals of Internal Medicine, 155(2), 97–107. Low health literacy is associated with more hospital admissions and poorer ability to take medicines correctly.
Open the source (opens in a new tab)Source 8
Stacey, D. et al. (2024). Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews, CD001431 (209 studies, 107,698 participants).
Open the source (opens in a new tab)Source 9
Latenstein, C.S.S. et al. (2020). The association between patients’ preferred treatment after the use of a patient decision aid and their choice of eventual treatment. Health Expectations, 23(3), 651–658. For gallstones and inguinal hernia, the preference after the decision aid matched the eventual treatment in 86% of cases.
Open the source (opens in a new tab)Source 10
Benker-Hasani, E. et al. (2026). Effect of previsit AI-generated document summaries on consultation time and patient–physician communication: the AI-INFOCARE randomized controlled trial. NEJM AI. 300 neurosurgery patients, Dresden: consultation 21% shorter (median 25 vs 33 minutes); interaction rated higher, especially on shared decision-making.
Open the source (opens in a new tab)Source 11
Kessels, R.P.C. (2003). Patients’ memory for medical information. Journal of the Royal Society of Medicine, 96(5), 219–222.
Open the source (opens in a new tab)Source 12
Dwyer, T. et al. (2023). Use of an artificial intelligence conversational agent (chatbot) for hip arthroscopy patients following surgery. Arthroscopy, Sports Medicine, and Rehabilitation, 5(2), e495–e505. 26 patients, no control group: the chatbot answered 31% of questions itself.
Open the source (opens in a new tab)For everyone who receives care, and everyone who gives it.
