The buzz around AI has been both energizing and overwhelming. Between the promises of 100x shipping speeds and existential worries about job security, it can be hard to find clarity. For product leaders, especially, the signal-to-noise ratio often feels off. What does it actually mean to use AI well in our daily workflows?

At Tint, we’ve stopped thinking about AI as a productivity boost and started thinking about it as a team member. That shift has been quietly transformational.

The Problem with “Speed First”

In product development, speed is often treated as the holy grail. But moving fast without clarity leads to poor outcomes. Documentation lags. Context gets lost. Decisions don’t get captured. And in an async, distributed environment, that can quickly snowball into misalignment.

We’ve seen this firsthand. At Tint, we intentionally keep our team small and nimble—by design. It gives us strategic advantages: we stay close to the customer, make decisions quickly, and avoid the overhead that can bog down larger teams. Our team is senior, capable, and focused on solving some of the most complex challenges in embedded insurance.

But that also means staying on top of every decision, every shift in context, and every update across multiple tools is a real challenge. And because we’re not in a position to cut corners—or risk the ambitious outcomes we want to achieve—we needed a way to ensure nothing falls through the cracks.

This is where AI started to become indispensable. We didn’t need another senior product leader—we needed a more junior skillset that could augment our existing team. Something (or someone) to extend the value of our senior talent by taking on the foundational tasks that keep us aligned and efficient.

Reframing AI as a Teammate

That’s where AI comes in—not as a replacement, but as a junior teammate. Someone to offload repetitive tasks, tighten up operational gaps, and surface insights that might otherwise go unnoticed.

We found it helpful reframe the discussion as an open role we were looking to fill and drafting the role, the tasks, and the way we’d measure performance. This ultimately led us to creating our new junior teammate with AI.

The Role

We asked ourselves, “If I could hire someone today, what would their role be?” The answer was clear: a junior PM or product assistant focused on documentation hygiene—capturing decisions, keeping records current, and ensuring no detail slips through.

The Tasks

Next, we asked ourselves, “What tasks would I trust them with?” We all felt comfortable starting with small but meaningful tasks like summarizing meetings, identifying decision threads in Slack, updating PRDs, and flagging when something is missing or unclear.

Measuring Performance

Finally, we had to ask ourselves, “How would I measure their performance?” For us that looked like measuring accuracy, completeness, and how often their work reduces the need for clarification or rework.

From there, we “assign” those jobs to AI, with the same level of oversight and feedback we’d give a new hire.

A Practical Example: Documentation

One of the most painful and persistent challenges in product workflows is keeping documentation current. In a typical week, decisions at Tint happen across Slack, Linear, Slite, and calls. Because we move fast, keeping PRDs up to date without assistance is daunting—but essential. And the consequences of falling behind aren’t just inconvenient—they pose a real risk to product quality. So we imagined what it would look like to bring on a junior PM solely to own documentation hygiene: capturing decisions, tracking context, and highlighting gaps.

That’s now one of our AI assistants. It still needs supervision, but it’s dramatically reduced the time we spend chasing updates—and increased our confidence in the documentation we do have. This shift has helped us ship faster and smarter.

Why This Matters for Embedded Insurance

Embedded insurance demands precision. The margin for error is low, and the cost of misalignment is high. You’re dealing with regulated products, complex workflows, and partnerships that rely on mutual trust. If your team doesn’t have a clear, shared understanding of what’s being built, it’s not just inefficient—it’s risky.

AI, used well, helps us reduce that risk. It gives us leverage: not just to move fast, but to move with clarity. That clarity is what allows Tint to be a reliable partner for brands, underwriters, and insurers alike.

What We’ve Learned

This process has taught us valuable lessons that we’ll take with us in each AI implementation we do.

• AI won’t replace your judgment—but it can extend your reach.
• Thinking of AI as a teammate unlocks new ways to delegate.
• The best use cases aren’t flashy; they’re foundational.

We’re still learning. But by treating AI as a core part of our product team—with the same expectations and accountability we’d give a new hire—we’re building a smarter, more resilient team.

And this isn’t just about efficiency today. By starting here—training our senior product team to work with, guide, and evaluate AI—we’re also building fluency for the future. This fluency strengthens our conversations about how AI integrates into our own product and enhances how we think about its broader potential across the embedded insurance ecosystem.

In embedded insurance, that kind of foresight isn’t just helpful—it’s essential.