Background:
I’m 10 years into building healthcare technology companies, previously the Founder and CEO of Biofourmis. I set out on a mission 10 years ago to harness the power of data and technology to bring the right care to every person, no matter where they are. My sole focus now is OutcomesAI. My ambition is to build this company with a 10-year view, spending my time and resources to build a safe AI for healthcare.
Our Mission
Amplify care capacity through Artificial Medical Intelligence (AMI).
The Company
I believe that providing safe, accessible, engaging, and personalized healthcare is the moral priority of our time. The most meaningful impact will come from technologies that enable delivery of high-quality healthcare at-scale. In the coming ages we will see great advancements in general-purpose Artificial Intelligence (AI) performing at expert-level. By contributing to the early-stage development, we can set the course for a safe medical AI future for humanity.
Hence the goal of OutcomesAI: to develop specialized and safe Artificial Medical Intelligence (AMI) to supercharge healthcare teams, and provide engaging and personalized care to patients, all for better clinical outcomes.
Our company journey will take time and will require a championship team dedicated to the mission, deep clinical collaborations, and AI engineering innovation to achieve a mass-market impact and adoption. We will face risks, safety concerns, and adoption challenges. However, if we succeed, we will make a positive impact on healthcare and build the largest medical AI company on the planet.
The Present
Today, we continue to see an unprecedented labor shortage in healthcare. The world will be short of 13 million nurses by 2030, and another 300k specialists are needed in the U.S. alone. In addition, our aging population will make it increasingly difficult for healthcare organizations to cater to patient needs, further intensifying capacity constraints. As a result, we foresee rising patient safety concerns, delaying care, decreasing patient satisfaction, and increasing cost of care.
If we want more care capacity, we need more productivity – and this means more automation.
The Possibility
AI is about to boost productivity in all sectors and create a new industry (projected to be $110B by 2030) by seamlessly enhancing human with machine capabilities. Zooming out, we are in the early stage of a revolution that is headed for a world with Artificial General Intelligence (AGI) – a reality where AI can think, learn, reason, and interact with the physical world when embedded within humanoids. AI will be capable of performing tasks better than humans.
Today, nurses make up nearly 50% of the healthcare workforce. However, 70% of them reported being over-skilled for the roles they were performing in their day-to-day work. The same is true for doctors, if not worse. Considering the costly and lengthy education programmes required for doctors and nurses, this job-skill mismatch represents a huge waste in human capital. I believe building safe medical AI that allows clinicians to dedicate their time to the most valuable and rewarding tasks will unleash clinical workforce productivity and boost care quality. Economic return will be a natural byproduct.
Our first step will focus on building AI companions that are deeply integrated into clinical workflows to assist nurses and care teams to be highly productive and focus more time on direct patient care. Think of it as a wingman for care teams that automates over 50% of their workload ranging from administrative tasks, triaging, documentation, to clinical decision support, and even patient engagement.
Over time, nurses could shift their attention to highly complex cases as AI becomes capable of interacting with patients to coordinate care and support routine, non-diagnostic tasks. For most of the time, nurses could even leave the loop altogether – driving the cost of care down even further. This advanced automation will boost clinical productivity in unprecedented ways.
I also believe with the advancements in robotics, especially humanoids, will expand machine capabilities to tackle highly complex tasks that can only be done by humans today. Looking ahead, we aspire to integrate our medical AI into humanoids to care for the ~1 billion elderly population in the comfort of their homes. We will not place the AI in roles that make diagnostic and treatment decisions, which could inflict harm on humans. Instead, our focus is on providing resources for jobs that care teams don’t want to and shouldn’t be performing.
The Solution
Thankfully, we are still in the early stage of the current AI revolution, which presents us with a unique opportunity.
There are two schools of thought on how to develop real-world medical AI: identify AI-solvable tasks and implement bespoke solutions or reverse the process and build a general medical intelligence that fits into every step of the patient journey. I believe the time is ripe for us to pursue the latter approach. Instead of building occasional road signs, let’s build the Google map for healthcare – navigating patient care journeys by accurate understanding of the entire landscape, status, and the destination – and over time, we will have full self-driving on most roads.
At OutcomesAI, we believe building a specialty-focused clinical operating system that addresses the entire patient care journey is the best route to unleash the power of artificial medical intelligence (AMI) and will have the largest overall impact.
How We Can Do It
In 10 years of studying and building companies, I’ve never seen a potential market size like what medical AI can bring. Arriving there will require significant advancements in technology. Today, even building AI companions with advanced medical reasoning capabilities to assist nurses and clinicians is still incredibly challenging, let alone a fully integrated system that works across the entire patient care journey. We’re heads-down and focused on making substantive leaps in areas that are necessary to put medical AI into clinicians’ and patients’ hands. They include:
- Artificial Medical Intelligence (AMI): Unlike other models which are built for singular data types – our team is focused on building a multi-modal foundation model (using text, audio, images, sensor signals) for healthcare applications – ensuring safety and clinical performance by pre-training on trusted clinical data and fine-tuning using human feedback.
- Care Delivery Operating System:We aim to augment the entire patient journey from diagnosis, treatment to follow-up care, handling administrative and clinical tasks and providing timely clinical decision support for clinicians. A fully workflow-integrated system is required on which clinical tasks operate and information flows. To achieve this goal, we distill best practices and efficient clinical workflows, integrate up-to-date evidence and guidelines, and redesign processes for personalized care, all into a single operating system powered by our AI, providing unified and seamless user experience.
- Safety: It’s essential that our AI will be able to assist care teams safely. We are ensuring safety by extensively validating performance against clinical benchmarks, and having it tested by the broad clinical community.
- Return on Investment (ROI): We foresee not only needing to deliver a high-quality product, but to deliver measurable ROI. We are building tools to measure ROI across various touchpoints to matter to our partners, be it incremental revenue, cost savings or improved patient engagement.
- Foundation for AI Nurses and Physical AI: Building a fully autonomous AI Patient Navigator (AI-PNs) and embodied AI requires extensive training, huge amounts of data and a community of partners. We are using this phase to lay the foundation to build a safe, highly skilled, and multilingual AI nurse to assist patients with their care needs, first through screen and eventually at bedside.
How We Can Do It
In summary here is the first phase of our Master Plan:
01 BUILD ARTIFICIAL MEDICAL INTELLIGENCE
02 BUILD SPECIALTY CARE DELIVERY OPERATING SYSTEM
03 INTEGRATE THE OS INTO EXISTING CLINICAL WORKFLOWS
We have the potential to alter the course of history and fundamentally improve the lives of millions of clinicians, nurses, and patients.
It’s time to build.
Kuldeep Singh Rajput
Founder & CEO
