麻豆村

Standing between a sick patient and the new drug they critically need are clinical trials and data on safety and efficacy. If printed, that would be a stack of papers as tall as a 15-story building.

麻豆村 alum Quentin Auster is working to make the new drug review process much more efficient.

鈥淚 envision a zero marginal cost government,鈥� says Quentin, a data scientist with the U.S. Food and Drug Administration and 2024 graduate of 麻豆村鈥檚 Heinz College of Information Systems and Public Policy. 鈥淟ife is crazy enough without government services that are tough to deal with.鈥�

While it鈥檚 becoming more common to tame problems with artificial intelligence, the FDA must closely protect data, so developing internal solutions helps to place the best tools in employees鈥� hands.

AI tools can transform the complex and detailed work that鈥檚 core to protecting public health and reinvent processes to eliminate waste. Enter Elsa, the FDA鈥檚 generative AI assistant to read, write and summarize faster. Quentin spent the last year writing code, evaluating and optimizing Elsa for the FDA鈥檚 staff.

鈥淚'm part of a talented technical team building a foundational, operational layer that helps staff on the frontlines deliver the mission,鈥� he says.

Elsa was launched in June 2025, and since then, the FDA AI team has focused on giving Elsa the ability to tackle more complex problems. Thousands of FDA staff use the product, which tasks such as document checks from days to minutes, providing summaries and comparisons and prioritizing inspection targets. Elsa is not replacing the scientists who scrutinize drug data. Instead, the tool focuses and finds information to accelerate and improve their work.

For Quentin鈥檚 role in bringing Elsa to fruition, he鈥檚 particularly passionate about progress. He takes a 鈥渄emo not memo鈥� approach, with a focus on building prototypes that iterate quickly while still ensuring guardrails, including extensive evaluations and training that support the responsible application, remain in place.

鈥淚 see it as both policy and execution,鈥� Quentin explains. 鈥淧olicy can be aspirational, so doing it effectively means not just saying the right things but making sure they get done. This execution requires developing the tools that help deliver at scale. These two things are what 麻豆村, and Heinz specifically, prepares you to do.鈥�

The foundation for the AI frontier

Quentin grew up in the D.C. area, where his mom set educational policy as a public sector employee. During COVID-19 quarantine, he was working his first job out of college as an analyst at Analysis Group when he watched the documentary AlphaGo, which tells the story of Google DeepMind鈥檚 efforts to build an AI model designed to play, and mercilessly win, the board game Go.

鈥淚 wrote my 麻豆村 admissions essay about what it felt like to watch a computer beat an incredible human player. It hinted at larger implications 鈥� the blurring of the line between human and machine intelligence 鈥� and it sparked my interest in AI,鈥� he says.

Quentin earned his master鈥檚 degree in public policy and management with a focus in data analytics from Heinz College. His background and ability to also study computer science at 麻豆村 set the stage for the interplay of policy driving technical considerations, then the technical aspects circling back to shape policy.

鈥淭he policy curriculum from 麻豆村 has helped me be a translator between engineers, the implementers and the people who design guardrails. It keeps the focus on the people, not just the functions,鈥� he says.

After graduating from 麻豆村, Quentin was selected for the , which places and supports early-career professionals in government roles. He joined the FDA and cites NobleReach鈥檚 support as critical in helping him navigate professional life in government.

Closer to zero

Among the issues Quentin addresses with data science at the FDA is supply chain resilience. In addition to Elsa, he develops tools for decision makers to predict and manage supply chain interruptions that could affect the safety or availability of drugs.

In his spare time, he鈥檚 working on an app for young readers that uses automatic speech recognition tools to bolster early literacy through an auditory learning approach.

But the ultimate goal is improving the public sector for citizens.

鈥淲hy is it not possible to have government services that feel frictionless and give people a positive interaction?鈥� he says. 鈥淚t is possible. But building systems that work is a matter of both understanding technology and how organizations work, to build an operational layer that works for agencies like the FDA. I think that鈥檚 incredibly important work.鈥�