At AWS re:Invent 2025, we often talk about "speed," but in the world of Life Sciences, speed literally saves lives.
In a fascinating interview on the expo floor, IQVIA (often called "the largest company you’ve never heard of") sat down with NVIDIA and AWS to discuss how they are dismantling the barriers to drug discovery.
If you are interested in how Agentic AI and GPU acceleration are reshaping healthcare, here is your 5-minute summary of the collaboration that is changing the game.
The Billion-Dollar Problem 💸
Lucas Glass, SVP of Technology at IQVIA, laid out the staggering reality of the pharmaceutical industry:
- Time: It takes 10 to 15 years to bring a new drug to market.
- Cost: The price tag sits between $1 billion and $2 billion…
At AWS re:Invent 2025, we often talk about "speed," but in the world of Life Sciences, speed literally saves lives.
In a fascinating interview on the expo floor, IQVIA (often called "the largest company you’ve never heard of") sat down with NVIDIA and AWS to discuss how they are dismantling the barriers to drug discovery.
If you are interested in how Agentic AI and GPU acceleration are reshaping healthcare, here is your 5-minute summary of the collaboration that is changing the game.
The Billion-Dollar Problem 💸
Lucas Glass, SVP of Technology at IQVIA, laid out the staggering reality of the pharmaceutical industry:
- Time: It takes 10 to 15 years to bring a new drug to market.
- Cost: The price tag sits between $1 billion and $2 billion.
- Risk: The success rate is only about 10%.
The goal of this partnership is simple but ambitious: Shrink these numbers.
The "Triumvirate" of Innovation 🤝
The session highlighted a three-way partnership where each giant focuses on their "comparative advantage" to solve this problem.
1. IQVIA: The Domain Expert 🧠
IQVIA runs a massive portion of the world’s clinical trials and acts as the largest healthcare data broker. They possess the data and the scientific know-how but realized they shouldn’t be in the business of building data centers.
- The Strategy: Use their massive historical data to design better clinical trial protocols (experimental designs) faster, predicting outcomes before a physical trial even begins.
2. NVIDIA: The Optimization Engine 🏎️
Lindy Wu from NVIDIA explained that their role goes beyond just selling hardware. Their goal is actually to lower the cost of compute.
- The Tech: They provide the GPU-accelerated software stack.
- The Key Innovation: NIMs (NVIDIA Inference Microservices). These are containerized, pre-optimized AI models that are "enterprise-ready." They allow IQVIA to take complex models and run them anywhere—specifically on AWS—without needing to manually optimize for the hardware.
3. AWS: The Scale & Infrastructure ☁️
Matt Carr from AWS highlighted the concept of "Undifferentiated Heavy Lifting."
- The Role: AWS provides the global infrastructure and scale. IQVIA doesn’t need to be an expert in HVAC, electrical engineering, or server maintenance.
- The Integration: AWS provides the platform (like Amazon Bedrock or Amazon EC2) where NVIDIA’s NIMs and IQVIA’s data meet. This allows IQVIA to spin up massive compute power for a simulation and spin it down instantly to save costs.
Key Technical Takeaway: The Rise of "Agentic AI" 🤖
The buzzword of the session was Agentic AI.
We are moving past simple chatbots. In Life Sciences, "Agents" act as digital teammates.
- Administrative Relief: A huge portion of healthcare costs is administrative (claims, denials, appeals). AI Agents can handle these repetitive workflows autonomously.
- Clinical Design: Agents can analyze thousands of previous clinical trial protocols to suggest the optimal design for a new drug, turning a process that takes months into days.
"We don’t want to take the human out of the loop... but how many administrators do you really want in a hospital system?" — Lucas Glass, IQVIA
Why This Matters for Developers 🛠️
Even if you aren’t in healthcare, the architectural pattern here is a lesson for all builders:
- Don’t Reinvent the Wheel: IQVIA uses AWS so they don’t have to build servers.
- Optimize at the Edge: They use NVIDIA NIMs to ensure their code runs as efficiently as possible.
- Containerize Everything: The ability to move models easily via containers (NIMs) allows for flexibility in deployment.
📚 Learn More & Get Started
Inspired to build your own Agentic workflows or explore Life Sciences on AWS? Check out these resources:
- NVIDIA NIMs on AWS: Learn how to deploy NVIDIA Inference Microservices on Amazon SageMaker and Bedrock. [Read the Docs](#)
- AWS for Life Sciences: Explore how AWS is powering the next generation of biology and clinical trials. [Explore the Hub](#)
- Amazon Bedrock Agents: Start building your own autonomous agents today. [Get Started](#)
Did you catch the interview? What do you think about the shift from "Generative AI" to "Agentic AI"? Let us know in the comments!