Nexmedis: The GPT of Healthcare

A few years ago, I was misdiagnosed multiple times and mistreated in Indonesia. Only after I went to Singapore that I was properly diagnosed with stage 3B cancer.

Unfortunately, I was not alone.

Millions of Indonesians seek medical treatment abroad, causing us to lose $11.5Bn annually. The Indonesian health system lacks the resources to properly serve 280 million populations, leading to suboptimal clinical services.

Indonesia as an archipelagic country has a challenging geography that hinders the country from achieving equitable health services.

Indonesia suffers from a shortage and a lack of distribution of health workers. Indonesia's doctor-per-population ratio is only 0.6:1000, lower than WHO's standard of 1:1000 and so much lower than neighboring countries like Singapore and Malaysia. Even then, research showed that 54.47% of them are located in Java and Bali. According to the Indonesian Medical Council's data, there are fewer than 50,000 specialists with a whopping 70% located in Java island.

Health workers also experience massive administrative burdens. Up to 22% of doctors' time is spent on administrative tasks and paperwork. This is due to the low rate of digital adoption within health facilities. Based on the Ministry of Health's data, as of April 2024, only 28% of all facilities (including hospitals, clinics, and GPs) have compliant electronic medical records. The majority of them are located in areas with decent internet connections. 

Being severely overworked, doctors are unable to provide ample time for patients to consult. Consultation time is a WHO-established measure of healthcare quality. Based on our research and backed by Cambridge research, In Indonesia, the average consultation length is only 3+ minutes. This affects the doctor-patient relationship and the opportunity to dig deeper into the patients' conditions. As a result, the patient's clinical outcomes are not optimal.

[Product demo here]

Nexmedis provides an AI-powered electronic medical record (EMR) to streamline health facilities’ operations, advance doctors' clinical services, and improve patient's clinical outcomes. It is offline-capable and available in both web and mobile applications, allowing health workers in areas with unstable internet connections and low resources to get digitalized. 

Our value proposition is the AI called Clinical Decision Support (CDS). It analyzes patients' real-time and historical health data and generates 5 probable diagnoses with their ICD-10 codes in < 2 seconds. 

By utilizing our AI, doctors can have a virtual assistant that helps streamline the consultation process and reduce the likelihood of medical errors without disrupting existing workflows. Our LLM Model functions as a Clinical Decision Support tool by analyzing data sources from the electronic medical record (EMR), including:

• Subjective data, i.e. patients' complaints, pain scales, and chronic illnesses
• Objective data, i.e. vital signs and physical examination notes

It then generates suggestions for five probable diagnoses based on identified patterns, accompanied by corresponding ICD-10 codes. It helps doctors make more informed decisions and reduces the time, cost, and possibility of medical errors. Evidence-based, patient-specific diagnosis recommendations improve patient's clinical outcomes.

For hospitals that already have an established electronic medical record, we can serve them through API-based Clinical Decision Support.

Despite being launched in August 2023, we have successfully digitalized 300+ health facilities and 4,000+ users across the country. This is a much faster acquisition compared to the industry benchmark of 65 health facilities per year. Because of our AI, we have won awards and recognition from key stakeholders, such as the Ministry of Health, East Ventures, JETRO, Siloam Hospitals Group, and Tech in Asia. We recently got appointed as the official partner of the Ministry of Communications and Information Technology for healthcare digitalization, particularly in remote areas.

Our solution serves all kinds of health facilities, from GP practices, clinics, to hospitals. The users vary from health workers (doctors, nurses, midwives, pharmacists) to non-health workers (admins, cashiers). They are located across the nation, even in areas with a lack of electricity and internet connections.

The majority of our target population still works with manual, paper-based documents, including patients' medical records, medical inventory, and financial transactions. It leads to inefficient clinical services and difficulty in making a comprehensive decision. 

Our solution changes the way they work, from paper-based to electronic. It streamlines the operations, allows management to make a strategic decision, and helps health workers advance their services. As a result, healthcare quality offered to patients is improving significantly.

Our team has complementary skill sets, from business, technical, to medical, with years of experience to build an impactful solution. The founders have proximity to the communities and issues we are solving.

Our CEO, Yehuda, has led nationwide digital transformation in many institutions. He understands what it takes to drive a successful change in transitioning from manual operations to digitalized ones. Additionally, he has conducted health-related research since he was young and was able to win a cancer-focused national research competition. This is why he can design the system with a strong technology behind it.

Our Chief Medical Officer, dr. Almer, is a doctor and clinic manager who has served underprivileged communities for years. He understands what the health workers need to help them serve the communities optimally. He has reviewed multiple EMR solutions and found a massive gap between them. He helps design and iterate the products to ensure it is the best product the users can have and that it closes the gap between what the market needs and what is available.

Our COO, Matilda, is a cancer survivor who has experience seeking medical treatments in Indonesia, Singapore, and the US. She has pitched the solution to many fellow patients and survivors with positive feedback since they believe that technology can help their doctors make better decisions for them. She also lived in a remote area, so she helped design a product that can facilitate health workers in low low-resource environment.

In conclusion, we develop Nexmedis with the communities at the center of our decision and design. Our strong understanding of the communities, combined with their involvement and input, allows us to deliver impactful solutions.

We have launched our AI-powered EMR since August 2023 and gained traction that is more and faster than the industry benchmark. By March 2024, we serve 300+ health facilities, 4,000+ health workers, and 200,000+ patients across the nation. More than 90% of the health workers use our web app.

For Advanced AI, we are finalizing contracts with major hospitals in Indonesia.

We are applying to Solve because our mission to help Indonesia achieve equitable health services aligns with MIT Solve's. To do that, we hope to tap into MIT Solve's network and resources to accelerate our growth and impact.

We believe MIT Solve is an amazing opportunity for us to connect with like-minded people and institutions whom we can further learn and collaborate with. Looking into the Solvers' and applicants' profiles, we are excited to have deeper discussions on how we can improve our solutions and tackle the challenges that they might have faced.

We are excited to gain mentorship and strategic advice that can help us scale up and measure impacts moving forward. We are also keen to explore new opportunities that can elevate our presence and impact in the region. 

If chosen as a Solver, we believe the funding can boost our efforts to reach more underprivileged communities across Indonesia and provide assistance to our users with low resources.

Nexmedis is uniquely positioned to lead the Indonesian market because we combine our innovative AI, called Clinical Decision Support, with the electronic medical record (EMR). We incorporated real-time and historical data to develop a scoring system integrated with our EMR system. We have the largest, localized datasets that we use to train our AI so that it understands the local context; not only the diseases but also the medications that take into account the patients' allergies and drug-to-drug interactions.

We collaborated with several medical schools in Indonesia to validate our AI model and gain a deeper understanding of local contexts, prompts, terminologies, and other relevant factors. It can help doctors in low-resource settings to provide more optimum clinical services.

Our direct and indirect competitors even express interest in using our AI in their EMR because they see the value in using advanced technology that is an extension of the doctors' knowledge.

Because of our Clinical Decision Support, the two ministries chose us to become a forefront startup in digital health.

The Ministries can connect Nexmedis with local governments or institutions in challenging areas where other providers are hesitant to serve. It helps the Ministry of Health accelerate the digitalization.

Health facilities, even in the smallest unit, use our AI-powered electronic medical records to securely track, store, and analyze large volumes of patients' data. Our seamless medical ecosystem fosters inter-division collaboration and optimizes the workflow, even if the facilities face personnel shortages.

Clinicians can utilize our Clinical Decision Support as their virtual assistant that rapidly analyzes patients' historical and real-time data and delivers 5 probable diagnoses in <2 seconds. The recommendation takes into account the latest medical advancements, local context, and endemic diseases. It acts as an extension of clinicians' knowledge. Combined with the EMR, it will reduce the likelihood of medical errors without disrupting existing workflows.

Patients as beneficiaries can benefit from the overall improved quality of services and experience better clinical outcomes. Patients will receive recommendations for personalized medications and treatments that take into account their allergies, conditions, and drug-to-drug interactions.

Our goal is to make high-quality healthcare available to all. We equipped health workers with cutting-edge technology to help them streamline operations, advance clinical services, and improve patient's clinical outcomes.

We use the indicators below to measure our progress:

• Number of healthcare facilities, users (health and non-health workers), and beneficiaries (patients)
• Area coverage, particularly those in remote and low-resource regions
• Applications usage and turnaround time compared to the one pre-Nexmedis/pre-digitalization
• Performance and accuracy of diagnosis recommendations, accompanied by the corresponding ICD-10 (diagnosis) code; we cut the process from 3 days to 3 seconds
• Feedback from users and completion time

For our electronic medical record, we developed a web and mobile application that is offline-capable to ensure users who live in areas with connectivity issues and low resources can utilize it. For the Clinical Decision Support, we developed and refined our LLM model through fine-tuning and established a knowledge base from it.

Steps:

• We gathered a vast array of medical datasets from various sources
• We train our AI models to identify patterns and correlations among symptoms, causes, diagnoses, treatments, and probable diagnoses
• We incorporated real-time data to develop a scoring system integrated with our Electronic Medical Record (EMR) system, widely utilized across multiple regions in Indonesia
• We collaborated with several medical schools in Indonesia to validate our AI model and gain a deeper understanding of local contexts, prompts, terminologies, and other relevant factors

8 FT staff, 4 PT staff, 2 interns

6 years of research and ~1 year since incorporated

At Nexmedis, we strongly believe that diverse backgrounds, upbringings, and perspectives are crucial in creating human-centered innovation. One in three founders is a woman with a triple minority. Our team is a combination of individuals from minority gender, ethnic, religious, and disability groups. We provide flexible working arrangements that accommodate diverse circumstances and locations.

We make sure everyone has their voice heard regardless of their position and background. We provide equal opportunity for leadership positions and encourage everyone, even the most junior staff, to contribute to the discussion.

More than 90% of our users are female and we make sure they have a safe space to ask us if they have any challenges in using our product. We encourage them to provide feedback to ensure that our solution keeps improving and beneficial for them.

Our business model is B2B modular SaaS. This means the customer can subscribe to our AI product that meets their needs. We offer two solutions:

• AI-powered EMR
  • End-to-end management system with AI
  • Suitable for health facilities with:
    • Manual medical record
    • Outdated electronic medical record
  • Typically used by: GP/nurse/midwife practices, clinics, community health services, and small hospitals
• Advanced AI
  • Plug-and-play, API-based Clinical Decision Support
  • Suitable for health facilities with:
    • Established, localized electronic medical records
    • Telehealth services
  • Typically used by: large hospitals

The subscription period is typically 1-2 years and many of them have expressed interest in using our solutions for the longer term (>5 years).

They are excited to use our products because we are the first one to provide a clinical decision support system that understands the local context and prompt. Indonesia is a tropical, archipelagic country with a unique disease profile. Some areas are endemic of malaria, for example, and some illnesses like leprosy are still prevalent. This nuance is something our technology can capture, thus increasing the accuracy of the recommendation.

Our plans are as follows:

• Convert pre-contract agreements into signed contracts for both products (currently hundreds of health facilities are in the pipeline)
• Raise funds from venture capitals and grants, particularly those focusing on health tech and AI, to further accelerate our expansion
• Partner with esteemed institutions for research and product distribution
• Launch our new AI products in the next few months (already receiving interest from major hospitals)
• Ensure we operate as a lean team with high productivity, thus keeping our operations cost-effective

We have secured $250,000+ of contract value for our AI-powered EMR with hundreds more facilities to be onboarded in the next 6 months.

Additionally, we are currently partnering with esteemed institutions such as Universitas Gadjah Mada and Universitas Surabaya for joint AI/biomedical engineering research. We also have joint operations with multiple institutions to further expand our target market.

While we are currently bootstrapping, we have secured $20,000 in prize money after winning competitions (e.g. Ministry of Health, Tech in Asia) and received perks worth $300,000 to further support our operations.