In Conversation with Forian’s Ben Miller: Applied AI, Predictive Models, and the Future of Agentic Systems

In this episode of A Tale of Two AIs, Razi sits down with Ben Miller, Director of Applied AI at Forian. Ben brings over 15 years of experience building AI systems that generate real business value, spanning healthcare, aerospace, media, and technology. Ben’s career uniquely combines deep scientific rigor with hands-on execution, and his to AI has always been driven by curiosity and problem-solving.

We caught up with Ben to talk about what it takes to operationalize AI, how to structure teams for success, and where agentic AI fits into the future of enterprise decision-making.

Applied AI: Beyond Models to Decisions

For Ben, applied AI is less about building models and more about influencing decisions and delivering measurable value. 

He points to examples from his career, such as a pilot churn model at United Technologies. By predicting which employees were at risk of leaving, his team implemented targeted interventions (i.e. hybrid work schedules, flexible hours) that ultimately saved 120 jobs in the first year, translating to an estimated $5 million in savings.

Ben stresses that the Applied AI function hinges on experiments: “Something that isn’t Applied AI is where you’re spending all this time and you never even get to the starting line. You never get something deployed. Even if it’s a garbage model, deploy it and get that feedback loop. Start building an understanding of where you’re making mistakes.”

Building Teams for the AI Era

When it comes to scaling AI teams, Ben emphasizes adaptability, problem-solving, and engineering mindset over deep theoretical expertise. “I lean towards folks who are experimental, curious, and willing to take risks,” he says.

He also stresses the importance of using the right tools and tech stacks to accelerate results while maintaining guardrails for quality and governance. 

The Role of Agentic AI

Ben is already experimenting with agentic AI in his day-to-day work. He uses agents for tasks like bug finding, quality control, and data validation—micro projects where agents act as specialized assistants. “Agents can do things like go click around an application, find bugs, document them, create tickets…They’re small, focused, but immensely helpful,” he says.

Looking ahead, Ben sees a network of specialized agents, each expert in a specific area, working together to amplify intelligence across an organization. “I think of it as a network of agents feeding into a larger system, much like a team of humans,” he explains.

Predictive AI Remains Core

While agentic AI is gaining attention, Ben underscores that structured predictive AI still drives the majority of enterprise value. “Predictive models give you deterministic outcomes that can be acted on with confidence,” he says. He envisions agents leveraging the outputs of these models to provide context, interpret results, and recommend interventions—creating a more intelligent, decision-focused system.

Looking Ahead

For Ben, the next frontier is combining predictive and agentic AI to drive actionable insights in real time. Teams that embrace experimentation, leverage the right tech, and focus on measurable outcomes are the ones poised to succeed. 

Listen to the full episode to hear more of Ben’s insights on applied AI, building effective teams, and harnessing agentic AI in the enterprise.