Why MedNotes? AKA How Did I Get Here?
Hello to all my fellow healthcare professionals, tech enthusiasts, and change-makers.
Throughout my career, I've walked two distinct yet interconnected paths—one rooted in the operational challenges of running multinational clinical trials, the other in the operational challenges of delivering care to patients around the world. Both paths, founded in the hope that technology might allow these services to scale and deliver more without breaking the dedicated individuals who operate at the coalface. These dual tracks have not only shaped my professional life but have also led me to co-found MedNotes, a synthesis of all I've learned and hoped to achieve in healthcare.
From the bustling hospitals of Melbourne to the boardrooms of global tech enterprises, my journey began in the late 1990s after graduating from the University of Melbourne. I was fortunate to complete my medical degree alongside a science degree in physics and mathematics, which kept me connected to the field of computers at the dawn of the internet. As a physician trainee, I was thrust into the heart of medical practice, witnessing firsthand the impact of direct patient care. In parallel, I honed my skills in Perl, HTML, SQL and (later) Javascript, exploring the early potential of PalmPilots and dial-up internet connections. Armed with this early tech experience, it quickly became apparent that the healthcare sector, though noble and critical, was riddled with inefficiencies that burdened clinicians and healthcare workers. Similarly, the global clinical trial processes we relied on as doctors for the delivery of better medicines were also plagued by inefficiencies, leading to delayed development of new treatments or prohibitively high costs. Addressing these inefficiencies in both sectors, through technology, presented a tremendous opportunity for improvement.
I embarked on two distinct yet complementary paths early in my career. The first was co-founding MedSeed, a company that developed a decision support architecture to connect with lab information systems. This system provided desktop and PalmPilot-based guidance to doctors, notably in antibiotic stewardship—the foundation of a system that remains in use today. However, despite its innovative approach, MedSeed struggled to gain critical mass due to the nascent IT infrastructure in healthcare and the limited adoption of handheld computers at the time. Consequently, I sold the company and moved on. The second path led me to co-found Health Research Solutions, which specialized in large-scale data capture for drug safety studies and real-world evidence. This venture was more successful, eventually being sold to a U.S. multinational. This sale propelled me into the heart of the drug development, tech, and capital boom in the U.S., where I spent nearly 15 years. My experiences there culminated in the co-founding of DrugDev, a multinational tech company dedicated to building systems that address the operational challenges of managing multinational clinical trial projects. Throughout this journey, my growing expertise in drug development technology contrasted with a persistent desire to translate these innovations back into the healthcare system.
When I sold DrugDev in 2017, I was determined to return to improving clinical practice. Many of my peers, mid-career consultants by then, shared harrowing stories from the front lines: long waiting lists, extensive administrative burdens, underfunded clinics, and crumbling infrastructure. What struck me most was the stark reality that EMR systems, once promising during my MedSeed days, had paradoxically increased their workload through their ubiquity. These systems demanded more clinicians to jump through more hoops with less staff support, leading to more errors and forced adaptations of messy, real-world medical practice into rigid, predefined workflows. Communication breakdowns were common, with staff employing numerous workarounds just to keep things moving.
This led me to co-found Induction Healthcare in the UK, centered around a straightforward app that allowed doctors to quickly find phone numbers within their healthcare system - saving time by bypassing the switchboard and dramatically easing communication delays. This user-driven approach saw significant uptake—doctors were clearly still open to technology that tangibly eased their burdens, despite their reservations about technology overall. We expanded the platform to enable patients to digitally connect with their hospitals and to engage in video consultations with their doctors. During the COVID lockdowns, particularly in the UK where Induction Healthcare now supports over 85% of the hospital market, our products proved indispensable. This experience reaffirmed that both doctors and hospitals were willing to adopt low-friction tech solutions that integrate with existing EMR systems and infrastructure with minimal information redundancy.
Despite the big wins achieved during this period, the journey still highlighted to me the persistent burden of paperwork still plaguing doctors in primary and specialist care across diverse markets such as the US, Canada, the UK, and Australia. While we've managed to overcome significant hurdles—like implementing widespread use of handheld computers in healthcare (smart phones!), facilitating remote patient-doctor interactions (WebRTC, chat and pub-sub technology), and building supportive IT infrastructure (EMR systems)—the relentless paperwork, often uncompensated, remains a formidable challenge.
So it was with great surprise and pleasure that I began experimenting with OpenAI's GPT playground in early 2022. The ability of AI to mimic human-like interactions and automate routine tasks was instantly promising. One of my initial experiments was to see if the system could draft a medical referral letter. The potential was undeniable. Yet, it was clear that significant work lay ahead. GPT-3, while impressive, had its limitations. No doctor would feel comfortable uploading sensitive patient information to a non-transparent system. The language it used just wasn’t the way a professional doctor would write. The hallucinations were risky. Moreover, learning about the questionable methods of training these AI systems—such as potentially bypassing copyright laws or scraping content—did little to inspire confidence. Athon Millane and I recognized the need to develop our own models, to provide a platform for doctors and healthcare systems to train these models securely, to handle patient deidentification effectively, and to ensure the user experience added no additional friction to their already demanding roles.
The elegance of the MedNotes system lies in its simplicity—from the user's perspective. A new doctor can begin using it within seconds, without prior training, and immediately appreciate its value—the kind of "wow" factor that feels almost magical, yet is backed by complex, optimized AI working quietly in the background. This includes advanced functions, nuanced training, and specialized models tailored to produce healthcare-specific outputs. Above all, it ensures patient confidentiality and aims to reduce, not increase, medico-legal risks.
In our vision, MedNotes is the missing piece of the puzzle. It is designed to be low friction, adaptable as an adjunct to telehealth or in-person care, compatible with any EMR system, and accessible whether the doctor or healthcare is on their phone, in the clinic, or at the bedside. Our tools are crafted to give doctors and allied healthcare professionals extra minutes with their patients, ensure they can take a lunch break, help a non-English speaking patient feel understood, and allow healthcare workers to leave on time, confident that their administrative tasks are completed and their patients well cared for.
We still have a long journey ahead. We aim to encourage doctors to integrate tools like MedNotes into their workflows, release more features to aid the community, and unveil additional tools from our labs that can streamline billing, reduce missed medical opportunities, and enhance patient engagement.
These AI-driven tools are on the horizon. But this week, we celebrate the launch of our MedNotes Foundation Model, a testament to the considerable effort of our team.
Thank you for being part of this adventure. Your support, insights, and passion for innovation are what drive us forward.
Dr. Hugo Stephenson
CEO & Co-Founder, MedNotes
Why MedNotes? AKA How Did I Get Here?
Hello to all my fellow healthcare professionals, tech enthusiasts, and change-makers.
Throughout my career, I've walked two distinct yet interconnected paths—one rooted in the operational challenges of running multinational clinical trials, the other in the operational challenges of delivering care to patients around the world. Both paths, founded in the hope that technology might allow these services to scale and deliver more without breaking the dedicated individuals who operate at the coalface. These dual tracks have not only shaped my professional life but have also led me to co-found MedNotes, a synthesis of all I've learned and hoped to achieve in healthcare.
From the bustling hospitals of Melbourne to the boardrooms of global tech enterprises, my journey began in the late 1990s after graduating from the University of Melbourne. I was fortunate to complete my medical degree alongside a science degree in physics and mathematics, which kept me connected to the field of computers at the dawn of the internet. As a physician trainee, I was thrust into the heart of medical practice, witnessing firsthand the impact of direct patient care. In parallel, I honed my skills in Perl, HTML, SQL and (later) Javascript, exploring the early potential of PalmPilots and dial-up internet connections. Armed with this early tech experience, it quickly became apparent that the healthcare sector, though noble and critical, was riddled with inefficiencies that burdened clinicians and healthcare workers. Similarly, the global clinical trial processes we relied on as doctors for the delivery of better medicines were also plagued by inefficiencies, leading to delayed development of new treatments or prohibitively high costs. Addressing these inefficiencies in both sectors, through technology, presented a tremendous opportunity for improvement.
I embarked on two distinct yet complementary paths early in my career. The first was co-founding MedSeed, a company that developed a decision support architecture to connect with lab information systems. This system provided desktop and PalmPilot-based guidance to doctors, notably in antibiotic stewardship—the foundation of a system that remains in use today. However, despite its innovative approach, MedSeed struggled to gain critical mass due to the nascent IT infrastructure in healthcare and the limited adoption of handheld computers at the time. Consequently, I sold the company and moved on. The second path led me to co-found Health Research Solutions, which specialized in large-scale data capture for drug safety studies and real-world evidence. This venture was more successful, eventually being sold to a U.S. multinational. This sale propelled me into the heart of the drug development, tech, and capital boom in the U.S., where I spent nearly 15 years. My experiences there culminated in the co-founding of DrugDev, a multinational tech company dedicated to building systems that address the operational challenges of managing multinational clinical trial projects. Throughout this journey, my growing expertise in drug development technology contrasted with a persistent desire to translate these innovations back into the healthcare system.
When I sold DrugDev in 2017, I was determined to return to improving clinical practice. Many of my peers, mid-career consultants by then, shared harrowing stories from the front lines: long waiting lists, extensive administrative burdens, underfunded clinics, and crumbling infrastructure. What struck me most was the stark reality that EMR systems, once promising during my MedSeed days, had paradoxically increased their workload through their ubiquity. These systems demanded more clinicians to jump through more hoops with less staff support, leading to more errors and forced adaptations of messy, real-world medical practice into rigid, predefined workflows. Communication breakdowns were common, with staff employing numerous workarounds just to keep things moving.
This led me to co-found Induction Healthcare in the UK, centered around a straightforward app that allowed doctors to quickly find phone numbers within their healthcare system - saving time by bypassing the switchboard and dramatically easing communication delays. This user-driven approach saw significant uptake—doctors were clearly still open to technology that tangibly eased their burdens, despite their reservations about technology overall. We expanded the platform to enable patients to digitally connect with their hospitals and to engage in video consultations with their doctors. During the COVID lockdowns, particularly in the UK where Induction Healthcare now supports over 85% of the hospital market, our products proved indispensable. This experience reaffirmed that both doctors and hospitals were willing to adopt low-friction tech solutions that integrate with existing EMR systems and infrastructure with minimal information redundancy.
Despite the big wins achieved during this period, the journey still highlighted to me the persistent burden of paperwork still plaguing doctors in primary and specialist care across diverse markets such as the US, Canada, the UK, and Australia. While we've managed to overcome significant hurdles—like implementing widespread use of handheld computers in healthcare (smart phones!), facilitating remote patient-doctor interactions (WebRTC, chat and pub-sub technology), and building supportive IT infrastructure (EMR systems)—the relentless paperwork, often uncompensated, remains a formidable challenge.
So it was with great surprise and pleasure that I began experimenting with OpenAI's GPT playground in early 2022. The ability of AI to mimic human-like interactions and automate routine tasks was instantly promising. One of my initial experiments was to see if the system could draft a medical referral letter. The potential was undeniable. Yet, it was clear that significant work lay ahead. GPT-3, while impressive, had its limitations. No doctor would feel comfortable uploading sensitive patient information to a non-transparent system. The language it used just wasn’t the way a professional doctor would write. The hallucinations were risky. Moreover, learning about the questionable methods of training these AI systems—such as potentially bypassing copyright laws or scraping content—did little to inspire confidence. Athon Millane and I recognized the need to develop our own models, to provide a platform for doctors and healthcare systems to train these models securely, to handle patient deidentification effectively, and to ensure the user experience added no additional friction to their already demanding roles.
The elegance of the MedNotes system lies in its simplicity—from the user's perspective. A new doctor can begin using it within seconds, without prior training, and immediately appreciate its value—the kind of "wow" factor that feels almost magical, yet is backed by complex, optimized AI working quietly in the background. This includes advanced functions, nuanced training, and specialized models tailored to produce healthcare-specific outputs. Above all, it ensures patient confidentiality and aims to reduce, not increase, medico-legal risks.
In our vision, MedNotes is the missing piece of the puzzle. It is designed to be low friction, adaptable as an adjunct to telehealth or in-person care, compatible with any EMR system, and accessible whether the doctor or healthcare is on their phone, in the clinic, or at the bedside. Our tools are crafted to give doctors and allied healthcare professionals extra minutes with their patients, ensure they can take a lunch break, help a non-English speaking patient feel understood, and allow healthcare workers to leave on time, confident that their administrative tasks are completed and their patients well cared for.
We still have a long journey ahead. We aim to encourage doctors to integrate tools like MedNotes into their workflows, release more features to aid the community, and unveil additional tools from our labs that can streamline billing, reduce missed medical opportunities, and enhance patient engagement.
These AI-driven tools are on the horizon. But this week, we celebrate the launch of our MedNotes Foundation Model, a testament to the considerable effort of our team.
Thank you for being part of this adventure. Your support, insights, and passion for innovation are what drive us forward.
Dr. Hugo Stephenson
CEO & Co-Founder, MedNotes
Why MedNotes? AKA How Did I Get Here?
Hello to all my fellow healthcare professionals, tech enthusiasts, and change-makers.
Throughout my career, I've walked two distinct yet interconnected paths—one rooted in the operational challenges of running multinational clinical trials, the other in the operational challenges of delivering care to patients around the world. Both paths, founded in the hope that technology might allow these services to scale and deliver more without breaking the dedicated individuals who operate at the coalface. These dual tracks have not only shaped my professional life but have also led me to co-found MedNotes, a synthesis of all I've learned and hoped to achieve in healthcare.
From the bustling hospitals of Melbourne to the boardrooms of global tech enterprises, my journey began in the late 1990s after graduating from the University of Melbourne. I was fortunate to complete my medical degree alongside a science degree in physics and mathematics, which kept me connected to the field of computers at the dawn of the internet. As a physician trainee, I was thrust into the heart of medical practice, witnessing firsthand the impact of direct patient care. In parallel, I honed my skills in Perl, HTML, SQL and (later) Javascript, exploring the early potential of PalmPilots and dial-up internet connections. Armed with this early tech experience, it quickly became apparent that the healthcare sector, though noble and critical, was riddled with inefficiencies that burdened clinicians and healthcare workers. Similarly, the global clinical trial processes we relied on as doctors for the delivery of better medicines were also plagued by inefficiencies, leading to delayed development of new treatments or prohibitively high costs. Addressing these inefficiencies in both sectors, through technology, presented a tremendous opportunity for improvement.
I embarked on two distinct yet complementary paths early in my career. The first was co-founding MedSeed, a company that developed a decision support architecture to connect with lab information systems. This system provided desktop and PalmPilot-based guidance to doctors, notably in antibiotic stewardship—the foundation of a system that remains in use today. However, despite its innovative approach, MedSeed struggled to gain critical mass due to the nascent IT infrastructure in healthcare and the limited adoption of handheld computers at the time. Consequently, I sold the company and moved on. The second path led me to co-found Health Research Solutions, which specialized in large-scale data capture for drug safety studies and real-world evidence. This venture was more successful, eventually being sold to a U.S. multinational. This sale propelled me into the heart of the drug development, tech, and capital boom in the U.S., where I spent nearly 15 years. My experiences there culminated in the co-founding of DrugDev, a multinational tech company dedicated to building systems that address the operational challenges of managing multinational clinical trial projects. Throughout this journey, my growing expertise in drug development technology contrasted with a persistent desire to translate these innovations back into the healthcare system.
When I sold DrugDev in 2017, I was determined to return to improving clinical practice. Many of my peers, mid-career consultants by then, shared harrowing stories from the front lines: long waiting lists, extensive administrative burdens, underfunded clinics, and crumbling infrastructure. What struck me most was the stark reality that EMR systems, once promising during my MedSeed days, had paradoxically increased their workload through their ubiquity. These systems demanded more clinicians to jump through more hoops with less staff support, leading to more errors and forced adaptations of messy, real-world medical practice into rigid, predefined workflows. Communication breakdowns were common, with staff employing numerous workarounds just to keep things moving.
This led me to co-found Induction Healthcare in the UK, centered around a straightforward app that allowed doctors to quickly find phone numbers within their healthcare system - saving time by bypassing the switchboard and dramatically easing communication delays. This user-driven approach saw significant uptake—doctors were clearly still open to technology that tangibly eased their burdens, despite their reservations about technology overall. We expanded the platform to enable patients to digitally connect with their hospitals and to engage in video consultations with their doctors. During the COVID lockdowns, particularly in the UK where Induction Healthcare now supports over 85% of the hospital market, our products proved indispensable. This experience reaffirmed that both doctors and hospitals were willing to adopt low-friction tech solutions that integrate with existing EMR systems and infrastructure with minimal information redundancy.
Despite the big wins achieved during this period, the journey still highlighted to me the persistent burden of paperwork still plaguing doctors in primary and specialist care across diverse markets such as the US, Canada, the UK, and Australia. While we've managed to overcome significant hurdles—like implementing widespread use of handheld computers in healthcare (smart phones!), facilitating remote patient-doctor interactions (WebRTC, chat and pub-sub technology), and building supportive IT infrastructure (EMR systems)—the relentless paperwork, often uncompensated, remains a formidable challenge.
So it was with great surprise and pleasure that I began experimenting with OpenAI's GPT playground in early 2022. The ability of AI to mimic human-like interactions and automate routine tasks was instantly promising. One of my initial experiments was to see if the system could draft a medical referral letter. The potential was undeniable. Yet, it was clear that significant work lay ahead. GPT-3, while impressive, had its limitations. No doctor would feel comfortable uploading sensitive patient information to a non-transparent system. The language it used just wasn’t the way a professional doctor would write. The hallucinations were risky. Moreover, learning about the questionable methods of training these AI systems—such as potentially bypassing copyright laws or scraping content—did little to inspire confidence. Athon Millane and I recognized the need to develop our own models, to provide a platform for doctors and healthcare systems to train these models securely, to handle patient deidentification effectively, and to ensure the user experience added no additional friction to their already demanding roles.
The elegance of the MedNotes system lies in its simplicity—from the user's perspective. A new doctor can begin using it within seconds, without prior training, and immediately appreciate its value—the kind of "wow" factor that feels almost magical, yet is backed by complex, optimized AI working quietly in the background. This includes advanced functions, nuanced training, and specialized models tailored to produce healthcare-specific outputs. Above all, it ensures patient confidentiality and aims to reduce, not increase, medico-legal risks.
In our vision, MedNotes is the missing piece of the puzzle. It is designed to be low friction, adaptable as an adjunct to telehealth or in-person care, compatible with any EMR system, and accessible whether the doctor or healthcare is on their phone, in the clinic, or at the bedside. Our tools are crafted to give doctors and allied healthcare professionals extra minutes with their patients, ensure they can take a lunch break, help a non-English speaking patient feel understood, and allow healthcare workers to leave on time, confident that their administrative tasks are completed and their patients well cared for.
We still have a long journey ahead. We aim to encourage doctors to integrate tools like MedNotes into their workflows, release more features to aid the community, and unveil additional tools from our labs that can streamline billing, reduce missed medical opportunities, and enhance patient engagement.
These AI-driven tools are on the horizon. But this week, we celebrate the launch of our MedNotes Foundation Model, a testament to the considerable effort of our team.
Thank you for being part of this adventure. Your support, insights, and passion for innovation are what drive us forward.
Dr. Hugo Stephenson
CEO & Co-Founder, MedNotes
Create clinical documentation
in seconds.
No credit card needed. Upgrade for unlimited letters, users & exclusive features.
Create clinical documentation in seconds.
No credit card needed. Upgrade for unlimited letters, users & exclusive features.
Create clinical documentation in seconds.
No credit card needed. Upgrade for unlimited letters, users and exclusive features.