Presentation & TLDR

This video was created as a substitute for the final presentation normally held in person at the end of each semester.

Background

The idea for this project arose when I had a really insightful talk with a medical student who was in her mandatory, unpaid one-year internship to gain experience in clinical work and patient care. Unfortunately, she told me, rather dissatisfied, that she barely got to actually work with patients because she spent up to six hours a day digging through paperwork and documenting procedures and consultations.

Curiously, this aligned closely with what I had heard from another doctor specializing in general medicine—long hours, heaps of bureaucratic paperwork, and poorly implemented digital systems that slow down processes rather than streamlining them.

This got me thinking: firstly, how might we clean up the messy, half-baked digitalization in healthcare so that doctors can finally focus on their patients again? And secondly, how might we simultaneously enhance the patient experience, which suffers due to time constraints and unappealing technology and user interfaces cluttering consulting rooms, leaving little space for meaningful doctor-patient relationships or safe spaces?

Survey

To gain some empirical insight on the prior assumptions, I conducted a survey among doctors and healthcare professionals about how they asses their situation and their satisfaction with the current state of digitalisation in their work routine.

TLDR:

The survey highlights the dual challenges of high administrative burdens and underutilized digital potential. Participants advocate for streamlined processes, robust digital tools, and infrastructural improvements to foster efficiency and job satisfaction in the healthcare system. These insights underscore the need for targeted digital transformation efforts to address current pain points.

For a more in-depth look at the survey as well as further research please scroll to the 'Research' Section.

Approach

Currently, many AI solutions tackle underlying and valid challenges by prioritizing ease of access, convenience, and personalization. This often manifests in mobile applications designed to remove the need for users to physically travel, engage with others, or navigate complex systems. With the assistance of AI, these solutions offer users an optimized, highly convenient experience. However, an increasing number of reports—such as those from the WHO (2025)—highlight a growing issue: social isolation. Alarmingly, up to one in four individuals now experience loneliness.

On the other hand, the motivation for becoming a doctor often stems from conviction, passion, and a desire to help others. This human connection is essential and is not easily replaced by technology.

Given this context, it would be counterproductive—and potentially harmful in the near future—to entirely replace traditional doctors with highly specialized AI systems that manage check-ups, medication, and therapy evaluations autonomously, even if they leverage advanced tools like digital-twin simulations. Such systems, while efficient, would lack the “human factor,” which is critical to the recovery process and plays a significant role in mental health.

In today’s healthcare landscape, meaningful patient interactions are often overshadowed by documentation and bureaucracy. Project NAVI was conceived to address this issue by creating space for genuine, human-centered interactions between doctors and patients, with AI serving as an unobtrusive assistant. Instead of placing AI at the center of the experience, NAVI emphasizes the importance of conversation. By streamlining consultations and minimizing unnecessary technological distractions, patients feel heard, and doctors experience less stress.

Concept

Instead of potentially driving people even further into isolation through hyper individualised and ego-centric experiences, Project NABI uses AI to deliver a highly personalised experience by bringing people together and making space for actual human connections.

This is achieved in three steps:

Preparation

Before the patient arrives, the doctor receives a comprehensive yet clear and concise briefing. This includes key information about the patient, such as their medical history, previous treatments and therapies, recent test results, current medications, the stated reason for the visit, and any relevant personal details from prior consultations.

The AI generates this overview by accessing and organizing patient records and data, presenting the physician with the most essential information in an easily digestible format. If the physician requires a deeper understanding of a specific topic, they can always explore further, with the option to manually search through the complete patient files at any time.

Consult

During this critical phase of the patient experience, perceptible technology is kept to a minimum and is designed to fade into the background entirely. Even the simple presence of a smartphone on a table can reduce focus and detract from the primary purpose – the conversation. To maintain a warm and inviting atmosphere, the room is free from screens, computers, keyboards, mice, and other cold or unaesthetic technology. Instead, the focal point is a tasteful wooden table, thoughtfully selected to suit the owner’s personal style (details on this are provided in the following section).

At this stage, the AI plays a supportive role by transcribing the conversation and gathering information on next steps, potential treatment options, and topics of interest. Additionally, the AI can provide visualizations in real-time, such as treatment pathways, test results, or medical illustrations. These visual aids are designed to support the physician’s explanations and facilitate collaborative decision-making.

Importantly, these AI suggestions are introduced with a subtle, unobtrusive glow, allowing the physician to decide whether or not to use them based on the context and their judgment of the situation.

De-Brief

After the patient leaves, the physician is provided with a concise summary of the consultation transcript, along with a draft of the agreed-upon treatment plan and any necessary additions, such as prescriptions, referrals, or upcoming appointments, ready for review. Once the physician signs off on the documentation, the AI automatically organizes and stores all new information appropriately. A processed overview is then sent to the patient, ensuring they have the opportunity to review and seek clarification if needed.

In this typically labor-intensive part of the process, the AI demonstrates its greatest potential by streamlining the physician’s workflow and significantly reducing the time spent on administrative tasks. For reasons of transparency and safety, the system allows manual review and override at any time, ensuring that professionals can address any irregularities or errors that may arise.

In Short

The AI supports physicians by providing concise patient briefings before appointments, assisting with real-time visualizations during consultations while always staying unobtrusive and optional, and streamlining post-visit tasks like summarizing transcripts and drafting treatment plans. It reduces administrative burdens while maintaining a high level of transparency and accuracy.

Interactive Table

To achieve a friendly and welcoming atmosphere and allow regular doctors offices to become a safe space all conventional technology and hardware is replaced by a tasteful and familiar boundary object – a wooden table.

Even though possible shapes and character are virtually infinite, there are a few characteristics that support face-to-face conversation.

In their research for the Ocular Series, Steelcase found this shape to be the most beneficial for teams working both in person and online at the same time. The distinguishing factor here is the seat arrangement that keeps a direct line of sight between participant while everyone is looking in the same direction – in this case a video-conference or TV-screen on the wall.

Showing typical desktop-setups within doctors offices.

Project NABI leans on this idea of the common goal being set through the seating arrangement. Instead of facing each other – doctor vs. patient – both parties face reality together. Ideally the table would be placed against a window to allow for natural light and the possibility for eyes to roam during the conversation.

I ultimately chose this shape because it combines a clear line of sight with an organic, inviting form that facilitates comfortable conversations between two people. At the same time, it remains practical and ergonomic for individual work. If needed, this shape also accommodates additional participants while maintaining an uninterrupted line of sight and excellent visibility of the content displayed on the tabletop.

The Projection

The ceiling-mounted projector adds a layer of interactivity to the table. It projects graphics, user interfaces, and full-color images directly onto the table’s surface. This system is complemented by a combination of cameras and LIDAR sensors, which enable precise and intuitive input from users.

On a side note: As an alternative to the outside-in approach, the table could incorporate an OLED panel and touch screen beneath a thin layer of veneer, allowing the table itself to display content. While this solution would be even more elegant and eliminate issues such as the projection illuminating hands or other obstacles, it would significantly increase the complexity of the product leading to higher R&D, production costs, and retail prices, making this option less practical. Additionally, such a design might lead to quicker obsolescence and make replacing the table much more expensive.

Video Prototype / Teaser

Research

TLDR:

The survey highlights the dual challenges of high administrative burdens and underutilized digital potential. Participants advocate for streamlined processes, robust digital tools, and infrastructural improvements to foster efficiency and job satisfaction in the healthcare system. These insights underscore the need for targeted digital transformation efforts to address current pain points.

Demographics and Professional Background:

• Respondents are primarily experienced healthcare professionals, with many having over 15 years of experience in various fields, such as general medicine, cardiology, oncology, and geriatrics.

• Positions range from nurses and medical assistants to freelance physicians and administrative roles.

Workload and Satisfaction:

• Patient interaction typically accounts for 3–5 hours daily, and satisfaction with this aspect is generally high, with most respondents reporting good focus during consultations.

• Administrative and organizational tasks take up a significant portion of the workday, with many spending 1–3 hours on documentation, management, and reporting, which is often described as burdensome.

Digitalization in the Workplace:

• The digitalization level varies, with practices using a range of tools like Turbomed, Doctolib, and Microsoft 365. However, full digitalization (e.g., paperless systems) is not common.

• Satisfaction with digital tools is mixed; technical issues and limited integration across systems hinder productivity.

Challenges:

• Frequent delays arise from coordination processes and administrative requirements.

• Many tasks currently handled by medical staff are perceived as suitable for delegation to support personnel.

• Compliance with regulations is reported as time-consuming and restrictive.

Aspirations for Improvement:

• Respondents seek better digital solutions, such as integrated platforms, AI-assisted scheduling, and tools for patient data management.

• Enhanced infrastructure, like stable Wi-Fi and mobile devices, is a common request.

• There is a desire for less bureaucratic workload, allowing more time for patient care and individualized treatment.

During the research phase of this project, I reviewed a wide range of academic papers and articles covering topics such as general UX design principles, user experience considerations specific to AI systems, the application of AI in healthcare, and other related subjects. This comprehensive exploration provided valuable insights and acted as a foundation throughout the project. A complete bibliography of the sources consulted is provided below.

Conclusion

To me this project once again shows the importance of carefully considered and human-centred UX Design. AI makes it very easy to think up products that supposedly address problems in dire need of sustainable solutions. However, the angle of approach is important here. AI should never be the solution in search of a problem. With its flexibility and and ability to variably respond to divers situations gives it the power to be the first technology ever to become actually human-centred instead of requiring people to adapt to it. However this takes courage and thought to find the essence of a certain situation, environment or system and extract the core experience we want to encourage. As a proof of concept Project NABI aims to take this step back, check to the left and right and filter out the doctor-patient experience instead of burying it beneath yet another chat interface.

This documentation may be updated and refined by 06/2025