Case Study

Lenfest AI Fellow spotlight: Ryan Thayer

The Minnesota Star Tribune

October 19, 2026

This profile is part of the Lenfest AI Collaborative and Fellowship Program’s 2026 Impact Report.

As an AI Fellow at The Minnesota Star Tribune, Ryan Thayer is exploring how artificial intelligence can unlock the value of decades of archived journalism and reshape how journalists search for, discover, and use news. We spoke with Ryan about the lessons he’s learned so far, what excites him most about AI, and where he believes the industry is headed next.

About his role and the problem he is trying to solve

Ryan’s path to the Star Tribune runs through software engineering, enterprise architecture, data science, and machine learning. When large language models and tools like ChatGPT arrived, he was well positioned to jump in.

At the Star Tribune, Ryan describes his role as part builder, part enabler. “My team and I have built AI tools from scratch. I help people from across the Star Tribune get started on their projects. I’ve helped others deploy their projects and apply engineering best practices if they’re not familiar with that.”

One of the projects he has focused on as an AI Fellow is archive search — helping journalists make sense of the vast trove of past coverage in ways that go far beyond simple keyword queries.

Rather than just answering narrow factual questions, the team is experimenting with richer, more conceptual queries.

“Instead of just a simple question-answer, it can be like a summary type of thing. Like, ‘Tell me what the Star Tribune has published about this wide-ranging topic.’”

The vision is to support both reporters and, perhaps eventually, the public. The system is currently in the proof-of-concept stage.

What has surprised him most during the fellowship

Ryan came into the fellowship used to working independently or within a single organization’s silo. One of the biggest surprises has been the level of cross-industry collaboration among Lenfest fellows.

He points to conversations with peers at other organizations in the cohort, where they are working on similar projects. “They have different perspectives. They are able to share things they’ve learned because they are at a different stage of a project, and I’m able to share feedback I’ve received because we went about it a little differently.”

“Just connecting with other people and hearing their progress and how they go about things has really helped me out a lot.”

One thing the industry gets wrong about AI

Ryan considers it a mistake to let AI do writing that requires a human touch.

“I read a lot of articles written by AI that I feel are not good. To me, there are a lot of words but not a lot of insight.”

At the Strib, there is a clear line: AI is not there to replace reporters or generate stories for publication.

“The reporters and journalists are still going to write their articles, but we’re looking to see what we can do to make things easier for them.”

The focus is on using AI to automate tedious, time-consuming work, such as background research, summarization, and other repetitive tasks, so journalists can focus on reporting and storytelling.

He’s noticed that once people learn more about practical ways AI can be useful, their attitude toward AI experimentation & adoption changes.

One lesson from his work that other newsrooms could apply tomorrow

Ryan’s key learning is product teams need to have discipline around the impact of their work.

“There’s a lot of cool stuff we can do, but for the work we’re doing, how does it move metrics? What are our main KPIs, and then how does the work move those metrics? And if it doesn’t, then we should probably be focusing on something else,” he said.

Archive search is a good example. The team constantly asks: What difference does this really make? If they can’t articulate the impact, it’s a sign to rethink priorities.

Right now, they’re still experimenting with which KPIs archive search can meaningfully affect: deeper engagement, new ways of using legacy content, and surfacing stories that people can’t easily find via search.

How the Star Tribune built Culinary Compass

What he is paying attention to right now

Ryan and his team are particularly focused on agentic development, using AI agents to help build and maintain software faster.

“How can we build things faster? How can we use agents to do the coding for us? And then again, where’s that line – what do people still have to do, what do experts still have to do, versus what can an autonomous agent take care of?”

The team is thinking about accelerating how they build everything, while still keeping humans in the loop for oversight, judgment, and deployment decisions.

What he is most optimistic about

How AI is reshaping roles and lowering barriers to contribution across the organization. Ryan is seeing a world where non-engineers can meaningfully shape products.

“Once we get an agentic development system set up, then someone who isn’t a software engineer can say, ‘Here’s an issue I see’ or ‘Here’s an improvement I need to make,’ and then the agents can make that update or at least do an evaluation and suggest updates.”

Similarly, product managers and designers can move faster.

“A product manager can mock up a design pretty quickly and hand that to the design team, and the design team is not starting from scratch.”

Or someone without deep coding experience can build a functioning prototype, then pass it to engineers to harden and deploy properly. “Again, that’s saving weeks from the requirements gathering process.”

For Ryan, this blurring of traditional role boundaries is one of the most exciting shifts underway.

The Fast Four

A book, article, or newsletter you’d recommend?
“Thinking, Fast and Slow” by Daniel Kahneman.

What’s one thing you’re learning that has nothing to do with AI?
Ryan points to curiosity as his ongoing personal focus – intentionally staying open to new ideas and digging into new topics as a mindset in daily life.

What local journalism challenge deserves more attention?
Hyperlocal coverage and truly distinctive niches: “How do we fit in versus what everyone else is doing? What niche can we really focus on that no one else is doing?”

One word that describes the next year of AI in news?
Fast-paced.