Transforming Healthcare Outcomes with AI and Global Talent

Rapid AI and AI in Healthcare

Pari: So tell us a little bit about RapidAI. What does your company do and what is the role of AI in healthcare innovation?

Amit: We are fundamentally an AI company operating in the healthcare domain. While we’ve done extensive work in stroke management, our focus spans the entire neurovascular disease landscape, particularly critical and acute care scenarios where time is absolutely crucial. Our solutions cater to acute conditions like different types of strokes, brain aneurysms, pulmonary embolisms, and similar vascular diseases. RapidAI has seen remarkable success, with our technology deployed across more than 2,250 hospitals worldwide. We process a staggering 14,000 scans every single day, and this number continues to grow rapidly at a yearly rate of 30% based on scan volumes. To date, we’ve analyzed over 10 million scans already.

However, our real impact lies in identifying over a million patients requiring life-saving interventions through our work. We offer a comprehensive mobile app and platform to seamlessly deliver AI-driven insights and facilitate their utilization. The mobile app serves as a vital channel for clinicians to access our AI analysis, view patient images, and leverage workflow tools alongside the core AI capabilities.

Pari: In what you’re doing, what is the role of AI? Give us a little bit of use cases. And how you’ve been using it so far.

Amit: AI plays a truly pivotal role in our work as a deep clinical AI company. In healthcare, you typically find various AI tools focused on basic triage and notification. These might tell a clinician they suspect something, but that’s about it. Our algorithms go way beyond that. That’s why we call ourselves deep clinical AI. In stroke cases, for instance, by running all our algorithms on the Edge cloud platform in parallel, we can deliver results from an AI perspective within 6-7 minutes, directly to a clinician’s mobile phone, email, or even PACS system. There are many ways clinicians can receive these AI results.

The Role of AI in Innovation

Pari: RapidAI has been around for the last few years, but the AI-led innovation seems to have accelerated in the last 12 months. Is there something changed in terms of how you’re using some of these newer algorithms?

Amit: There’s a clear understanding now that AI has a major role in healthcare, especially for image analysis and diagnosing diseases. At recent conferences like RSNA, clinicians are voicing concerns that simple triage and notification from AI aren’t enough. They need more than just a suspected finding. Ideally, AI should pinpoint the problem’s location, quantify and characterize the disease, and even measure and localize it precisely within the anatomy.

This is crucial because these details take clinicians a lot of time to determine, especially when images aren’t clear. AI that sheds light on these aspects and provides accurate localization and characterization becomes incredibly impactful for clinicians.

Pari: So coming into the kind of new innovations, Generative AI is not going to give the same answer every single time. But in healthcare, you need to be a lot more data-deterministic on how these algorithms work. Now, are you able to leverage some capabilities of Gen AI into your traditional healthcare capability?

Amit: Yeah, very interesting question. Pari, we’ve been actively exploring Generative AI (Gen AI) but there are limitations. While Gen AI can’t be used for FDA-approved diagnostics yet because it can be inaccurate, it has other valuable applications.
For instance, Gen AI can power intelligent chatbots with vast knowledge bases for easy information access. Even more promising, Gen AI can generate synthetic data for training algorithms. This is crucial because real-world data is scarce. However, real-world data is still essential for validating these algorithms, especially for FDA approval.
Overall, Gen AI offers exciting possibilities in healthcare, but responsible development is key, especially for diagnostics.

Healthcare-related challenges for AI

Pari: In less regulated industries e-Commerce, media, it is a lot easier, right? What are other healthcare-specific challenges? We talked a little bit about it when we discussed the Gen AI piece, but are there other healthcare-specific challenges in adopting AI in innovation?

Amit: High accuracy is absolutely essential for AI to be useful in healthcare. Even if an algorithm seems pretty good with 90% accuracy, that can still translate to a significant number of false positives and negatives. For clinicians, that can mean missed diagnoses or unnecessary procedures, which can seriously reduce trust in the technology. Just highlighting suspicious areas simply isn’t enough.

To truly be helpful to busy clinicians, AI needs to go beyond basic suspicion. Ideally, it should pinpoint the exact location of a problem, quantify its severity, and even characterize its nature. This ability to provide a more comprehensive picture is what makes AI truly valuable. However, achieving this level of detail is very challenging. One major hurdle is the limited availability of real-world data. While a massive dataset might make achieving all these goals easier, such a thing doesn’t exist. And for rarer diseases, obtaining real-world data becomes even more difficult. This lack of data adds another layer of complexity to developing truly effective AI for healthcare.

Talent in Healthcare AI

Pari: Very interesting, that is quite challenging. Right now the job market is very hot for talent with AI skills. So how, how do you retain talent at this time, especially around healthcare, and a fast-growing startup like RapidAI?

Amit:

Building talent is crucial, but the healthcare industry has a real advantage: purpose.

At RapidAI, we see the potential impact every day. Every minute spent improving our algorithms could save a patient’s life. This sense of purpose is incredibly motivating for engineers. They see the direct impact of their work – thousands of lives potentially affected by a single algorithm update. It’s a truly inspiring space to be in.

Pari: And you have set up a tech hub- a GCC in India, how is that team able to help in driving the innovation, especially since a lot of this work you do is largely for the US market?

Amit: Building the right team is crucial for us at RapidAI, and that’s why location isn’t a primary concern. We might be a young company, but we invested heavily in Bangalore right from the start. We made it very clear from day one that where an engineer sits doesn’t matter – it’s about matching talent with the work that needs doing.
This commitment is clear when you look at the Bangalore team’s ownership of core projects. Over the past year and a half, they’ve done an amazing job – the Bangalore team developed almost 80-90% of the entire edge cloud platform. That’s a critical piece of technology, the foundation that orchestrates all our algorithms.
And Bangalore’s contribution goes beyond the core platform. We’re actively building our mobile application development team there because we see the immense potential in the talent pool. As we continue to grow our US team, Bangalore will play an even bigger role. This global approach lets us move fast and efficiently.
We’re a late-stage startup now, but we’ve already achieved significant success and built a strong base of clinical evidence. In every region we operate in, including Bangalore, robust clinical validation is a top priority for us.
The bottom line is, that we’re committed to fostering a global talent pool, and Bangalore is a vital part of our present and future success. We see the opportunity to do more and more there, and that investment will definitely continue.

Pari: Got it. So it’s interesting that you just look at them as a common pool of talent and which, and you have a set of workload you need to get done and you just get done with this talent. There’s no major separation in the quality that these teams work on. And when talent joins Rapid AI, what can they expect in terms of their learnability?

Amit: At RapidAI, we are operating at the cutting edge of technology across multiple fronts – be it working closely with regulatory bodies like the FDA, tackling extremely complex technical and clinical challenges, or pushing the boundaries of workflow tools. As a deep clinical AI company, everything we do has immense depth and complexity.
Moreover, we have a massive global deployment of 2,250 hospitals, where our products are used daily by clinicians. This allows us to get instantaneous feedback, continuously improve, and directly impact lives saved. Working at RapidAI is an unparalleled learning experience due to this scale and the advanced nature of our work.

The teams, including our Bangalore site, are continuously challenged and thrilled by the innovative work they do. For instance, the Edge Cloud platform developed in Bangalore is a state-of-the-art, high-performance AI platform deployed at a massive scale. It pushes boundaries in parallel processing, sustainability, security, and serviceability. Every aspect operates at the highest level, enabling our deep clinical capabilities.

Pari: My final question, Amit. It seemed like the AI innovation way was again started to accelerate, right? What are some of the AI-led healthcare use cases you look forward to in the next two to three years?

Amit: The true promise of AI in healthcare is to democratize access to high-quality care, irrespective of geography, economics, or hospital resources. AI has the potential to normalize clinical expertise globally, making it available even in the most remote areas as well as developed nations.
Moreover, AI enables early detection of diseases at the earliest stages of their evolution. This upstream effect significantly improves patient prognosis and reduces treatment costs dramatically. By detecting diseases sooner, we can make healthcare far more equitable and accessible to all. AI allows us to move disease identification and intervention upstream, leading to better outcomes and lower costs across the board. This is the transformative power AI can unlock in healthcare worldwide.

Pari: Thanks a lot, Amit. Thanks for sharing insights on how AI is used in healthcare. The key challenges that it could create and the opportunities it creates for high-quality talent where they’re able to work on things that can give them a higher order of purpose compared to other industries. And finally, how AI could truly make healthcare equitable for people across the world? All of these are very insightful. and thanks again for sharing your perspective.

Amit: Thank you, Pari.

Pari: Thank you for tuning into this episode. We’ll be back soon with another leader, another exciting episode. Till then, take care and stay curious.

*This is an edited version of the conversation