Our Blog | 44ai

When Innovation Arrives in Everyday Life: Haidy at Adullam Hospital

44ai Team — Thu, 04 Jun 2026 08:00:00 GMT

How Adullam Hospital brings AI-powered documentation into clinical routine: A practical field report and a sign of what’s possible when technology, processes, and people work together.

Sometimes innovation begins with a simple question: How can we relieve our staff in their clinical day-to-day work, simplify documentation processes, and at the same time further develop the quality of our reports? Adullam Hospitals and Care Centers Basel asked exactly this question and tested Haidy as the answer.

The result: a clear yes to continuation and concrete figures that show why.

A Pilot with a Deliberate Learning Character

Since February 2026, Adullam Hospital, together with the Innovation & Transformation and ICT teams as well as the management of Medical Cross-Sectional Services (MQD), has conducted a structured AI pilot. The pilot was deliberately designed to be low-threshold, resource-efficient, and iterative.

The AI-powered documentation was tested in occupational therapy, physiotherapy, nutrition therapy, and psychological services with use cases ranging from progress notes to patient conversations and structured reporting. The pilot phase started in February and was extended for nutrition therapy and physiotherapy until the end of April. During this time, user experiences were regularly collected and evaluated.

The focus was not simply on introducing a new technology. It was much more about learning together, experimenting, critically reflecting, and finding out where artificial intelligence can create real value in everyday hospital life and where processes or framework conditions need to be developed first.

The Results: Measurable and Qualitative

Haidy supports the creation of structured medical documentation from conversations. This is intended to better prepare existing information, simplify report creation, and reduce administrative burdens. The goal: more time for medicine, therapy, and direct work with patients.

The strongest effect was seen in nutrition therapy. Even during the pilot phase, nutrition therapists saved 30 to 80 minutes per person per week. For a team of six, this corresponds to a potential of 3 to 8 hours weekly depending on workload. With optimized use and full system integration, 11 to 12 hours per week were identified as a realistic target.

Per Person / Week

30-80 Min

Time Saved

Team (6 People) / Week

3-8 Hrs

Team Time Gain

With Optimized Use

11-12 Hrs

Realistic Potential

Source: Internal Pilot Report Adullam Hospital, Feb.–Apr. 2026 · Nutrition Counseling (6 staff members)

Even more important than the time savings, however, were the qualitative results. Documentation was streamlined, workflows were more consciously questioned, and the focus during patient conversations was strengthened. Less documentation burden means more presence — exactly what Haidy is made for.

Not Everything Was Easy — and That’s Exactly What Was Valuable

The pilot also showed: AI-powered documentation is not a self-runner. In some areas, the results were not yet stable enough for an immediate rollout. Post-processing effort, the still-missing integration into parts of the existing workflow, and the necessary behavioral change in daily practice take time and support.

But that’s precisely why the pilot was so valuable. It wasn’t just about testing a digital tool — it made ways of working visible. It sparked questions, enabled discussions, and helped to better understand where technology can truly support and where processes, routines, or framework conditions need to be developed first.

Haidy® Reporter Registered as a Medical Device: What's Behind It?

44ai Team — Tue, 19 May 2026 08:00:00 GMT

Developing software for medical environments is subject to strict regulatory requirements. For us at 44ai, these are not just an obligation but a strategic ambition, which is why certification under the EU Medical Device Regulation (EU MDR) is an integral part of our product strategy.

Regulatory requirements create clear quality standards and a structured framework for the continuous development of our products. Our goal is not only compliance but also active participation in shaping standards, to take a pioneering role in developing safe, reliable, and high-quality software solutions that deliver real value in everyday medical practice.

Ongoing Risk Assessment

Medical devices must undergo continuous risk assessment procedures as a core component. For the Haidy Reporter, potential risks affecting the software were systematically analyzed, and residual risks posed by the product were evaluated.

This covers both technical risks such as faulty data processing or system failures, as well as usage-related risks such as misunderstandings or misinterpretations of the user interface. For the Haidy Reporter, this meant that all processes, from audio recording to result presentation, were assessed in detail.

The goal was not only to identify risks but to minimize them through appropriate measures or eliminate them as completely as possible. These measures include clearly designed and user-friendly interfaces, comprehensive technical validation, and continuous system monitoring.

Extensive Technical Validation

For AI-powered software tools in particular, technical validation is an essential component of the technical documentation. Manufacturers are required to demonstrate under realistic conditions that the software functions as intended.

We therefore subjected the Haidy Reporter to extensive validation to ensure that the AI-powered documentation delivers reliable and consistent results. Among other things, the software’s results were analyzed for confabulations, inaccuracies, and omissions using complex datasets.

A central aspect of this process is traceability. Every step and every test result was documented to ensure the transparency and reproducibility of the tests performed.

Confabulations

Confabulations describe clinical content that appears in the AI-generated report but has no basis in the original source. This includes invented diagnoses, never-prescribed medications, examination findings that were never conducted, or events in the patient history that never occurred. Hallucinations represent the most clinically concerning type of error, as they introduce false information into the patient record and can potentially lead to inappropriate treatment decisions.

Inaccuracies

Inaccuracies refer to information that is present in both the source and the AI-generated note but has been distorted during the transcription or summarization process. Examples include incorrect dosages, swapped side indicators (e.g., "left knee" instead of "right knee"), incorrect dates, or misattributed symptoms. Inaccuracies can pose a high clinical risk, particularly when safety-critical details such as medication doses, allergies, or surgical sites are affected.

Omissions

Omissions refer to clinically relevant information that is present in the original source but missing from the AI-generated note. They are particularly undesirable because, unlike invented content that may stand out due to its unusualness, missing clinically relevant information can only be detected if the reviewing professional independently remembers the omitted detail or cross-checks the source again.

Clinical Evaluation and Post-Market Surveillance

Alongside the technical perspective, clinical evaluation is a decisive factor on the path to conformity under the EU MDR. It answers the question of whether the software provides a direct or, as in the case of Haidy Reporter, an indirect clinical benefit for patients.

As part of the clinical evaluation, we examined the extent to which the application improves documentation quality, reduces workload, and relieves medical professionals. Scientific evidence from existing literature, comparative studies, and results from studies conducted with the Haidy Reporter were used for a well-founded assessment.

Report quality assessment results for an equivalent dataset
(relative to Haidy Reporter, lower is worse)

It is important to emphasize that clinical evaluation is not a one-time process but one that is continuously carried out as part of post-market surveillance. The manufacturer’s responsibility therefore does not end with placing the product on the market; on the contrary, after market launch, data on the Haidy Reporter’s usage, performance, and potential incidents continue to be collected and continuously evaluated. This includes feedback from users, new clinical findings from the literature, incidents from other comparable products, technical log data, and automated analyses of safety-relevant events during ongoing operation.

Post-market surveillance ensures that potential problems are identified early, enabling targeted improvements and ensuring long-term product safety. This ensures that the software meets the highest quality requirements not only at the time of market placement but throughout its entire lifecycle.

Quality Through Consistent Commitment

The core areas of risk management, technical validation, clinical evaluation, and post-market surveillance illustrate how complex and comprehensive the path to conformity under the EU MDR is.

For us as a manufacturer, this means not only fulfilling regulatory requirements but also a clear commitment to quality and safety. Achieving medical device status is therefore far more than a formal step. It represents a central milestone on our journey to responsibly deploy innovative AI technology in healthcare.

How to Use Haidy: The Physician's Perspective

44ai Team — Tue, 07 Apr 2026 10:00:00 GMT

Medical documentation is one of the most time-intensive tasks in clinical practice. With Haidy, we are enabling a new approach: an AI scribe that can document a complete consultation.¹ Haidy is an intelligent system that structures the content of conversations and documents and transfers it directly into the patient information system.

Thanks to integration with leading hospital information systems (including KISIM/Cistec, EPIC, INES KIS, WinMed, Axenita, and Aeskulap), Haidy can link existing information from the patient record with the current consultation and directly create new reports or progress notes. Future extensions will also cover appointment coordination and communication with other care providers.

More Than Speech-to-Text: Content Understanding as the Key

The decisive difference from conventional dictation solutions lies in the content-level understanding of conversations and documents (information in the broadest sense) and in the ability to structure this directly. Haidy doesn’t just transcribe spoken words — it interprets their medical meaning in context and structures the information accordingly.

Report quality measurably increases when as much clinically relevant information as possible is verbalized during the consultation.²³

This can be specifically supported through the consistent application of established medical communication techniques (such as NURSE).⁴ By naming emotions or repeating symptoms, Haidy also captures non-verbal information and correctly categorizes the significance of a symptom.

Verbalizing examination findings directly during the examination enables their immediate capture. Important elements of the doctor-patient relationship, such as shared decision-making about further therapy or next steps, simultaneously help Haidy document these correctly.

Practical Tips for Use in Daily Practice

The following approaches are recommended for optimal use:

Combination of Conversation and Post-Processing:
Particularly efficient is the combination of a guided conversation with subsequent dictation of report elements that were not verbalized directly in front of the patient or only become available later (e.g., lab or radiology results as well as the final clinical assessment).

Processing Unstructured Documents:
Haidy also recognizes and processes scanned or unstructured documents, such as lab reports in PDF form, and correctly transfers them into structured documentation.

Communication Techniques as Documentation Aids:
Standardized communication structures such as NURSE also prove to be an effective method for optimizing Haidy’s documentation quality.

Limitations of the Technology: Critical Review Remains Essential

AI scribes show their strengths primarily with standard clinical presentations — i.e., symptom patterns that follow a typical disease course. With diffuse or atypical manifestations, however, difficulties can arise, particularly regarding the clear articulation of the system’s own knowledge boundaries.

A critical review of the generated report is therefore essential, especially in complex or unclear cases.

Additionally, it should be noted that medical assessment, diagnosis derivation, and clinical reasoning will remain the domain of humans for the foreseeable future.⁵⁶

From Scribe to Decision Support Tool

Haidy’s reasoning capabilities — its so-called clinical reasoning — are developing rapidly. We expect that AI scribes will increasingly be linked with decision support functions in the future, providing evidence-based support for clinical decision-making processes.

For physicians, this primarily means substantial relief from documentation and administrative tasks.⁷

This allows physicians to refocus on clinical skills — which will not be replaceable by AI tools for the foreseeable future.

References

Sanmark, E. et al. Impact of an AI medical scribe after 375,000 notes generated across care levels in a European health system. medRxiv (2025). https://doi.org/10.64898/2025.12.06.25341757 ↩
Greenwood JD, Matthews MR, Overgaard JD, Ohde JW. Maximizing Efficiency of Artificial Intelligence-Enabled Ambient Scribes in Outpatient Settings: A Pragmatic Approach to Structuring the Patient Appointment. Mayo Clin Proc Digit Health. 2026;4(1):100339. doi:10.1016/j.mcpdig.2026.100339 ↩
Draper TC et al. Clinical AI Scribes in primary care: accuracy, error severity and implications for clinical practice. BMJ Digital Health & AI. 2025;1:e000092. https://doi.org/10.1136/bmjdhai-2025-000092 ↩
Patterson J, Kovacs M, Lees C. Ambient Artificial Intelligence Scribes: A Pilot Survey of Perspectives on the Utility and Documentation Burden in Palliative Medicine. Healthcare. 2025;13(17):2118. https://doi.org/10.3390/healthcare13172118 ↩
McCoy LG et al. Assessment of Large Language Models in Clinical Reasoning: A Novel Benchmarking Study. NEJM AI. 2025;2(10). DOI:10.1056/AIdbp2500120 ↩
Kim J et al. Limitations of large language models in clinical problem-solving arising from inflexible reasoning. Sci Rep. 2025;15:39426. https://doi.org/10.1038/s41598-025-22940-0 ↩
Tierney AA et al. Ambient Artificial Intelligence Scribes to Alleviate the Burden of Clinical Documentation. NEJM Catal Innov Care Deliv. 2024;5(3). DOI:10.1056/CAT.23.0404 ↩

Lights, Camera, Haidy!

Nils Eyer — Wed, 04 Mar 2026 10:00:00 GMT

There’s that one moment on a film set when it suddenly goes completely quiet. The crew is in position, the lighting is right, and someone calls “Action!” For most of us at 44ai, this was an absolute first — and, honestly, a bit surreal.

A few weeks ago we wrapped up filming for our first two company videos: a company video and a product video for the Haidy Reporter. What started as a simple brief (“We need a product video!”) turned into one of the most fun, slightly chaotic, and satisfying projects we’ve tackled as a team so far.

Two Videos, Two Stories

From the start, we knew we wanted to shoot two different videos. The product video was the heart of the project and had a single job: to make something technical feel intuitive. We wanted everyone who had never heard of Haidy to watch the video and think: “Ah, so that’s how it helps me.” No jargon, no complicated AI concepts. Just a clear, entertaining video showing what the Haidy Reporter does in a real clinical moment.

The company video was meant to convey something else: trust. 44ai is young and turns 2 this April. But a lot has happened in the two years since 44ai was founded. Today, alongside many others, the Children’s Hospital Zurich, Balgrist University Hospital, UKBB, Adullam Hospital, Schulthess Clinic, concordiaMed, and Swiss Post count among our clients. Haidy is integrated into leading hospital information systems such as KISIM, EPIC, Ines KIS, WinMed, Axenita, and Aeskulap. And our team has grown to over 10 employees, with backgrounds from ETH Zurich, the University of St. Gallen (HSG), and leading Swiss hospitals. The company video was our chance to tell this story properly.

Hospital Corridors and Our Office in Olten

All scenes were shot at two locations. The majority of the scenes were filmed in a real hospital. These spaces have something you can’t replicate: the light, the atmosphere, the feeling that something meaningful happens here every day. This gave the videos an authenticity that a studio could never have provided.

For the more personal moments, like the interviews, we returned to our office in Olten. We wanted to proudly show the place where the ideas that Haidy emerges from are debated. Having Jannis and Stefan speak in front of the camera there felt right.

The Team Behind and in Front of the Camera

Early in the project, we knew we wouldn’t achieve the quality standard we wanted for the video without help. ETUM Films was not only responsible for filming and production but was also a creative partner in preparation. They challenged us to think carefully about what we wanted to say and helped us translate it visually.

On our side, several team members stepped in front of the camera — no small feat for a group of engineers and researchers. Jannis and Stefan both appear in the company video and tell the story of why we founded 44ai and where we’re heading. To bring the patient perspective to life, we also hired a professional actress.

And This Is Just the Beginning

Producing these videos was a great opportunity to pause and reflect on what 44ai is today and where we want to go. When you’re in the day-to-day operational grind, there’s little time to clearly articulate your own story. The shoot forced us to do exactly that. And the result speaks for itself.

Both videos are now available on our website, on YouTube, and on LinkedIn. We look forward to your feedback!

The Videos

📺 Watch the 44ai Company Video on YouTube

📺 Watch the Haidy Reporter Product Video on YouTube