Within the framework of the NVIDIA GTC, organizations are at a crucial crossroads in their exploration of artificial intelligence (AI). The question is no longer whether they should adopt generative AI, but how to transition from promising experiments to production systems that deliver tangible value to their businesses. Those who can solve this challenge swiftly will position themselves with a notable competitive advantage, and inspiring examples are emerging that demonstrate the potential of this technology.
A standout example is Hippocratic AI, which is developing AI-driven clinical assistants to support healthcare teams in a critical moment where the workload has increased exponentially for doctors and nurses. Recently, during a hurricane in Florida, their system was able to contact 100,000 patients in a single day to verify their medications and provide preventive health recommendations; a task of almost unimaginable complexity to perform manually. Hippocratic AI is not just creating chatbots, but is transforming the delivery of healthcare services on an unprecedented scale.
However, the successful implementation of AI requires more than just advanced models or state-of-the-art GPUs. Throughout my experience managing data journeys, I have learned that an organization’s most valuable resource is its specific expertise and the quality of its data. In my current role leading the data and AI market, I often hear customers outlining their needs in this transformation process: they seek infrastructures and services they can trust, with a focus on performance, cost-efficiency, security, and flexibility, capable of operating at scale. In this context, AWS has excelled in providing effective solutions to operationalize advanced technology.
Another sector where generative AI is having a significant impact is content creation. Adobe, a leader in transforming creative workflows for over forty years, has integrated generative AI into their product suite, assisting millions of creators in their processes. Their Vice President of Generative AI, Alexandru Costin, describes their AI infrastructure as an “AI highway,” enabling rapid iteration of models and seamless integration into their applications. The success of their generative Firefly models, included in tools like Photoshop, demonstrates the effectiveness of this approach.
ServiceNow, positioning itself as the AI platform for enterprise transformation, is quickly incorporating this technology to redefine key business processes. Their architecture combines high-performance storage with NVIDIA GPU clusters for training and uses NVIDIA’s Triton inference server for deployment in production, allowing them to focus on developing AI tailored to each domain.
In turn, Cisco is methodically transforming its telecommunications applications suite. By separating their AI models from the applications and migrating their language models to Amazon SageMaker, they have created an architectural environment that promotes more agile development and cost optimization.
Ultimately, what Hippocratic AI highlights is the critical need for rigorous and secure architectures in environments where the implications are of utmost importance. Their approach, which combines over 20 specialized models, allows them to handle thousands of patient interactions while ensuring clinical safety.
The collaboration between AWS and NVIDIA, which has evolved significantly over the years, continues to adapt to meet the demands of the generative AI era. Together, they are empowering innovators from different sectors to transform their industries and discover new applications of this technology that has the potential to change lives.
via: MiMub in Spanish
