Kaushik Sarkar Believes AI Can Transform Disease Forecasting and Prevention

The climate crisis is rapidly becoming a global health crisis. Rising temperatures and changing environments are redefining when, where, and how diseases spread. From heatwaves to floods, frequent extreme weather events are stretching health systems to their limits, exposing gaps in preparedness and putting hundreds of millions at risk.

Kaushik Sarkar knows that with better data and broader access to decision-making tools, the world can move from reacting to health crises to preventing them. He founded the Institute for Health Modeling and Climate Solutions, also referred to as IMACS, as a global center of excellence or centralized hub of expertise hosted by Malaria No More. IMACS provides shared infrastructure and AI-powered predictive technologies to make data at the nexus of climate and health accessible and usable at scale. With usable data, health workers around the world can track diseases, such as malaria, dengue, diarrhea, influenza, tuberculosis, and other climate-sensitive diseases, rapidly identify high-risk areas, and implement measures that protect millions of people.

IMACS is now expanding its foundational AI work to increase reach and integrate more effectively within different health systems. In this interview, Kaushik shares insights about the work ahead and his vision for AI at scale that will contribute to a healthier future for all.

How did your journey inspire you to explore AI for humanity?

Ironically, my path into AI began with medicine. Twenty years ago, as a physician-scientist, I discovered how hidden data patterns could explain why certain children suffered co-infections of HIV and Hepatitis C. Back then, there were no open-source libraries like TensorFlow that could help us create machine learning (ML) models to make quick sense of large amounts of data. We also lacked access to the Graphics Processing Units (GPUs) required to train the advanced AI models. But I had a clear conviction that data could save lives when harnessed correctly! That conviction evolved into a career dedicated to building systems that make data more usable at scale.

Several years ago, at Malaria No More, we developed an AI-enabled system to forecast malaria incidence and guide public health interventions. The system was powered by one of the first hybrid neural networks, which essentially merges brain-mimicking neural networks with techniques like ML. Our success revealed that AI, when engineered for scale and precision, can be one of the most powerful instruments of our time to address complex challenges at the climate-health nexus. The realization led us to build IMACS to address the growing problem of infectious diseases exacerbated by climate change. Now, in its fifth year, we are expanding IMACS' services through a Central Data and Analytics Hub to go even further. Its biggest impact will be harnessing global data to bring climate-health intelligence to the fingertips of decision-makers, enabling them to solve the entire range of climate-driven health problems. So, two decades on, it really feels like a full-circle moment for me, from treating one patient at a time to building AI solutions that can protect millions.

What is the Central Data and Analytics Hub, and how does it leverage AI to support health planning and disease forecasting?

Our Central Data and Analytics Hub (CDAH) is designed to be the growth engine for IMACS. By providing a shared data infrastructure, it scales the impact of AI so that countries can manage and predict health risks faster, cheaper, and more accurately. Today, AI in health is advancing at a rapid pace. But, in the rush to innovate, we are often falling back on fragmented insights. For example, an AI model might show that heavy rainfall drives dengue outbreaks, while another model shows a link between the same disease and drought elsewhere. Both are correct within their context. However, neither tells the whole story, nor do they provide enough intelligence to make decisions related to localized health policies, programs, supply chains, or manufacturing.

The objective of CDAH is to provide a more comprehensive picture of the data so that we can create interventions and solutions that work for different contexts. It will have a network of expert AI models to inform all those different decisions. To accelerate its adoption, we are also developing a shared marketplace for these tools, connectors to fast-track their integration into existing national infrastructures, a training platform to strengthen local capacity, and a sandbox to simulate and evaluate. But it isn’t just these components that will unlock specialized knowledge. Backed by the Patrick J. McGovern Foundation, we are building an AI Foundation Model using multi-domain data across health, environment, and population to power the entire stack. We are also preparing a training program for our first cohort of public health decision-makers. Within the next three years, our goal is to expand the use of CDAH to 20+ countries to make faster, data-backed health decisions.

What are some design principles that help ensure the seamless integration of new AI products and services within existing national health system infrastructure?

Most health information systems predate the AI era, and replacing them is cost-prohibitive. The most effective way to help countries leapfrog into adopting AI without imposing huge capital expenditure and training burdens is through incremental innovation. At IMACS, we have seen that AI solutions that fit into a public health system’s existing technology stack with near-zero friction have the potential to halve implementation time and cut ownership costs anywhere between 40 to 80%. So, our first design principle is interoperability– tailoring our solutions to local data systems and digital infrastructure. Second, we co-design the solutions with the Ministries of Health and provide training and advisory services to leave behind sustainable expertise. Third, we focus on human-centered integration, meaning we design AI solutions around the workflows of policymakers, epidemiologists, and health workers instead of forcing them to adapt to the technology. The result is a 3-4x faster adoption speed and long-term operational independence for the countries we serve.

What core values drive your unique vision for impact in an AI-driven future?

I believe that the future must be human-led. Humanity must not hand over the wheel to AI. At IMACS, we don’t design AI systems to replace human judgment. We design AI systems to help decision-makers see further, act sooner, and make more informed choices. In our space, trust is the growth multiplier. High-stakes decision support systems must first earn the trust of the decision-makers themselves before they can create an impact.

To build and maintain that trust, we prioritize three simple values to anchor our work: integrity of purpose- we develop only what serves humanity and strengthens human resilience; equity of access– our efforts are targeted toward protecting the most vulnerable; and transparency of understanding– no matter how advanced the model, the design has to be explainable. That’s the standard to which I hold myself and IMACS every day.

Which visionary leaders, philosophies, or movements give you hope for a more human-centered AI future?

I am deeply inspired by the philosophies of the former President of India, the late Dr. A. P. J. Abdul Kalam. He once stated that Einstein’s equation needs to reach everywhere. Applying that conviction to AI, we must ensure that AI benefits are accessible to all. I see that principle in action today as nations and organizations converge on AI governance and digital sovereignty. I am hopeful that this year’s UN resolution to create an independent international scientific panel will get us closer to a more human-centered and safer AI future. At the same time, leading philanthropic and development investors are treating human-centered AI as a scalable asset class. This mindset shift from traditional aid to catalytic investment in mission-critical technology is what will define the next decade of growth in this space.

What is your 7-word autobiography?

Kaushik means sage, so I strive forward!