From startups to industry-leading enterprises, there is an insatiable desire for organizations to experiment with the idea of how the world could work differently. They explore bold, new ways for society to operate and meet market demand by focusing their efforts on developing innovative products and services. With 78% of companies using AI in at least one business function,1 we’re now seeing AI adopted by organizations of every size and sector around the world. And startups that use AI have been reported to obtain funding 2.5 times more quickly than those without it.2
Organizations use AI as catalysts for change—taking risks, challenging the status quo, scaling niche ideas, and reshaping entire industries. In season 1 of the Catalyst series, the possibilities unfold through the lens of three startups using Microsoft Azure infrastructure and NVIDIA’s domain-specific models and accelerated computing technologies to advance science, health, and technology. From digital twins to drug discovery, meet the companies driving social and scientific progress through new methods that amplify human ingenuity and skill.
Pangaea Data helps doctors close care gaps with AI
Diagnosing and treating health conditions, whether rare or prevalent, can be difficult for doctors and patients because patient records can span hundreds of pages long and clinicians have limited time to thoroughly review them, resulting in information remaining buried within them and patients being missed. Helping clinicians identify optimal clinical pathways faster leads to better care and improved patient outcomes. Pangaea Data set out to close care gaps by applying AI configured on clinical guidelines to emulate the process clinicians use to review electronic health records. This helps to find untreated and under-treated patients across hard-to-diagnose conditions at the point of care, without disrupting clinicians’ existing workflows.
For example, the UK’s National Health Service (NHS) applied Pangaea’s platform to identify six times more untreated and under-treated patients with the hard-to-diagnose condition, cancer cachexia, which can help halve the cost of treatment. Similarly, a US health system deployed Pangaea’s platform to recognize $9 million in additional revenue each year by closing care gaps for only one condition.
Pangaea’s platform runs on Microsoft Azure‘s globally compliant and secure cloud infrastructure and is integrated with NVIDIA’s comprehensive AI stack. By combining high-performance GPU-accelerated computing with optimized software frameworks and tools, Pangaea’s platform delivers timely and dependable insights trusted by clinicians and health care providers. This helps ensure equitable patient care and helps improves outcomes in a financially sustainable and scalable manner.
You’ve got to make sure that information is delivered to the physician in a compliant manner. And this is where Azure does a phenomenal job. The fact that they have already done the groundwork to be compliant across all different countries and territories. It immediately instills that confidence that you are a known entity, a trusted entity, beyond just the technology.
–Vibhor Gupta, PhD, Founder and CEO of Pangaea Data
Basecamp Research connects biotechnology with biodiversity
Basecamp Research sees nature as the key to scientific discovery. They have built one of the world’s largest biological databases to support life science researchers and companies working on drug discovery, product development, and biological research. This database of more than 10 billion novel proteins includes valuable tools such as gene editing proteins that have been collected from diverse natural environments around the world. The dataset is then used to train AI models, like Evolutionary Scale Modeling (ESM), that can help identify new targets, design novel proteins, and predict metabolic pathways, supporting faster development of innovative medicines and biological products.
AI has enabled Basecamp Research to set new benchmarks in control, novelty, and efficiency. Increasing their database’s annual growth rate tenfold required scalable computing resources, advanced models, and extensive data. By using Microsoft Azure infrastructure to operate their models and employing NVIDIA’s BioNeMo framework for training, they significantly increased operational speed and achieved scalability.
We’ve been leveraging Azure for this quite a lot, leveraging Azure Kubernetes Services (AKS) and the container registry to orchestrate this and to basically massively scale these annotation pipelines. NVIDIA comes in at the next stage when it comes to utilizing these datasets for model training.
–Glen Gowers, Co-Founder of Basecamp Research
Global Objects digitizes physical artifacts across industries
Microsoft sees AI digital twins as a foundational technology with many applications that can help bridge the physical and digital worlds. Digital twins provide virtual replicas of physical assets and systems, whether it’s a piece of artwork in a gallery, a prop in a film, or a simulation that can be used to expedite the production process in manufacturing. Global Objects shared how they’re using advanced AI and 3D scanning technologies to create photorealistic digital twins of props and objects across different locations. The company aimed to address the challenge of producing high-quality photorealistic content and fill that gap by training AI to map physical environments in the highest fidelity dataset with precision.
By blending artistic skills with advanced technology, they were able to make an authentic digital version of novel objects in the cloud. To accomplish this, they gathered a large subset of data about random objects to help generate new content that accurately replicates the exact specifications of each object digitally. Digitization of the objects can be used for not just preserving art or making films, but also creating new products and services across gaming, robotics, and healthcare.
The most important thing around all of this is processing. The files, the size, the speed by which we’re operating all require the best systems, servers, operators, GPUs that are in the market today. The relationship with Microsoft and NVIDIA is so powerful not only for us learning internally on our own systems and operations but engaging with our clients and ensuring we always have uptime, ensuring that we can always deliver with the most speed, efficiency, and fidelity as possible.
–Jesse Redniss, Chief Strategy Officer of Global Objects
Become a catalyst for change with Microsoft Azure and NVIDIA
Microsoft and NVIDIA offer industry-leading development tools, frameworks, and pre-trained models that are designed to support developers, researchers, and business leaders alike. By combining Azure’s trusted cloud foundation with NVIDIA’s cutting-edge compute capabilities, companies can deploy AI faster, iterate smarter, and stay ahead of the competition. The examples you’ve seen today are catalysts for a more inclusive and intelligent future, driving progress across diverse fields and industries. Watch The Catalyst Series.
1 McKinsey, The state of AI: How organizations are rewiring to capture value, March 12, 2025.
2 Crunchbase, The state of startups in mid-2025 in 8 charts: Global funding and M&A surge as AI fervor continues, July 22, 2025.
The post Transforming scientific discovery with Microsoft Azure and NVIDIA appeared first on Microsoft Azure Blog.