GenAI for Emotional Finance: Sooner Founder on Trust and Tech

Rose paid off her house and built savings, avoiding the stock market because it felt too risky. Then a friend from the gym recommended a stock. She put $25,000 into it and lost it all. The story, recounted by Johannes Seemann in a recent interview, is not about bad advice. It is about trust. People follow a person they have a good relationship with, even when logic says otherwise.

Seemann, an IDEO alum and former design director at Wells Fargo, spent two years running ethnographic research inside the bank before founding Sooner, a fintech startup built around someone's relationship with money. The human insight came first. The delivery mechanism came later, when GenAI made it feasible to offer judgment-free guidance at scale, 24/7, without the friction of finding the right advisor.

Sooner starts with a psychographic profile, a conversational alternative to a personality test that builds toward what Seemann calls "intentions" rather than goals. Goals are binary, hit or miss. Intentions flex when life disrupts the plan. A $2,000 vet bill would break a goal-based savings target but gets absorbed by an intention like "take care of your dog, stay aligned with my values."

Users preferred push-to-talk over free-flowing AI conversation, Seemann said. It handed control back. The AI surfaces options and reasoning. The person makes the call.

Seemann identified three design problems specific to AI that his team confronts daily. They are not prompt-engineering fixes solved once.

A model trained on broad cultural data projects value judgments onto what people say, and those judgments shape every future recommendation. One user described himself as frugal. The model might treat that as a stable trait, even when the user's behavior contradicts it.

Models also want to resolve contradictions into a tidy narrative. A young man told Sooner he wanted to be his own man making his own decisions in New York, while his parents still covered part of his rent. Both statements are true. A model under pressure for coherence flattens one out.

The longer a conversation runs, the more a model's early understanding of someone fades, right when the person is trying to make a bigger decision that depends on that context. Seemann is candid that he does not yet know how much of the fix belongs to Sooner and how much the underlying models still need to solve.

Wells Fargo, where Seemann built the strategic design function that ran the original research, pursues its own AI initiatives in wealth management. The bank's stock carries an Alpha Score of 62, reflecting moderate sentiment from investors weighing the margin potential of AI-driven personalization against the trust risks Seemann identified. The score suggests the market sees upside but wants proof that AI can handle the emotional side of money without alienating users.

Sooner is in beta. Seemann says he plans to let the product speak for itself.