The Questions Nicole Junkermann Asks Every Founder About AI

What problem are you actually solving?

The first question sounds simple and is often the hardest to answer well. Many companies describe their technology before they describe the problem it solves. Nicole Junkermann prefers to start the other way around. What is the specific pain a customer feels today, how do they handle it now and why is the current approach not good enough? A clear answer reveals whether the company exists to serve a need or to showcase a capability.

Artificial intelligence makes this question more important, not less. A model can be impressive in the abstract and still be a solution in search of a problem. The strongest founders can name the customer, the workflow and the moment in that workflow where their product earns its place. They talk about the problem with more energy than they talk about the technology, because the problem is what they are really in business to solve.

When a founder cannot describe the problem crisply, it usually shows up later as confusion about the customer, the pricing and the product roadmap. Starting with the problem is a way of testing whether the rest of the plan has a solid foundation.

What data do you depend on, and do you have it?

Artificial intelligence runs on data, so Nicole Junkermann asks early about the data a company needs and the data it can realistically obtain. Some products depend on data that is hard to access, expensive to maintain or tangled in permissions. Others have a genuine advantage because they can gather data that competitors cannot, often by being close to a workflow that produces it naturally.

The question is not only about quantity. It is about relevance, quality and the right to use it. A founder who has thought carefully about data can explain where it comes from, how it stays current, how it is protected and how it improves the product over time. A founder who waves the question away usually has more work to do than the pitch suggests.

Data also shapes defensibility. A product that gets better as more customers use it can build a lead that is hard to copy. That is why this question often leads naturally to the next one, about what protects the company once the idea is no longer novel.

What is your advantage once the novelty fades?

Novelty is a poor moat. A clever use of artificial intelligence can attract attention, but attention invites competition, and capabilities that feel rare today often become common tomorrow. Nicole Junkermann wants to understand what protects a company when the underlying technology is widely available. The answer might be proprietary data, deep integration into a customer's operations, a trusted brand in a sensitive field or simply better execution sustained over time.

This question separates a feature from a business. Some impressive products are really features that a larger platform could add, while others solve a problem so specific and so well that customers would not want to switch. The founders who think clearly about defensibility tend to have a more honest view of their market and a more durable plan for growth.

It is also a question about discipline. Building a lasting advantage usually means doing unglamorous work for a long time: refining a workflow, earning trust, improving reliability. Nicole Junkermann respects founders who understand that the advantage is built slowly, because that understanding tends to predict how they will behave when the market gets crowded.

Where does human judgement stay in the loop?

A theme that runs through the AI Overview is that artificial intelligence is most useful when it works with human judgement rather than around it. So Nicole Junkermann asks founders where a person stays in the loop, especially when the product touches decisions that carry real consequences. The best answers show a clear sense of which choices the system can make on its own and which require a human to review, approve or override.

This matters for trust and for risk. A product that quietly automates a consequential decision without oversight can create problems that are hard to undo. A product that designs the human role thoughtfully, giving the reviewer enough context and authority to act, tends to earn more trust from customers in regulated or high stakes settings. Founders who have thought about this usually have a more mature view of how their product fits into a customer's operations.

The question also reveals values. A founder who treats human oversight as an obstacle to remove sees the world differently from one who treats it as a feature to design well. Nicole Junkermann pays attention to that difference, because it shapes how a company will handle the harder moments that every serious product eventually meets.

Why these questions, and why now

None of these questions is exotic. They are deliberately plain, because plain questions are the ones that reveal whether a founder has done the thinking. The problem, the data, the advantage and the role of human judgement together form a simple map of whether an artificial intelligence company is built to last. Asking them is a way of moving past the demonstration to the durable substance underneath.

Nicole Junkermann brings the same questions to her work as the founder of NJF Holdings and its venture capital arm, NJF Capital, where backing companies across artificial intelligence and deep technology depends on this kind of clarity. On the podcast the questions are a way to learn. In investing they are a way to choose. In both cases they reward founders who understand their own business well enough to answer simply, which is often the clearest sign that a company is worth watching.