Innovate safely, decide wisely

AI lets you ship in days. It also lets you build the wrong thing for six months before you realize it. I help founders test the right assumptions first — so the speed works for you, not against you.

Book a discovery call →

Trusted by

The Problem

Speed hides the risks

Most AI teams discover these risks after a painful pivot. By then, six months are gone. The risks were always there, they just weren't visible.

Cartoon-style explosion with the word 'BOOM!' in bold yellow letters on a red burst background.

User risk

"Wow, cool" isn't proof of a genuine need, trust, or willingness to pay. Enthusiasm isn't a reliable indicator. Behavior is.

Isometric three-story building labeled 'Business' with large windows and surrounding greenery.

Business risk

You proved you can build it. The economics, channels, and model costs are still untested, and they'll bite you at month nine.

3D illustration of Earth showing continents North and South America, Europe, and Africa with two small white clouds.

Environment risk

The context your product depends on is moving — and most teams haven't mapped what they're exposed to.

Golden balance scale with a rock on the left side weighing more than the empty right side.

Bias

You keep building in the wrong direction, because no one has named the biases driving the roadmap.

The Method

Structured validation.
Not gut feel

Most teams skip from idea to build with assumptions still invisible. This method makes them explicit — so you can test the ones that matter most before committing resources to the wrong direction.

Six people sitting around a conference table with laptops and sticky notes, smiling and participating in a Discovery Sprint meeting in a bright office with plants on the windowsill.

1. Map your assumptions

Every belief your product depends on gets surfaced, about users, business model, and market context. Most teams find 20–40 they've never said out loud. You rank them by evidence gap and consequence.

Man and woman smiling while looking at two computer screens, one showing a business model canvas and the other a video conference call.

2. Design the right experiment

For the riskiest assumption, you design the minimum test needed to learn, not the minimum to ship. A prototype that isolates one variable. An interview that doesn't lead. A test that could actually prove you wrong.

Five colleagues in a meeting room with a wooden table, presenting information on a wall-mounted screen and a paper board.

3. Make a defensible decision

Evidence collected. Assumptions updated. The team aligns on what it means. Build, pivot, or kill — with the reasoning documented so the decision holds up six months later, not just in the room.

The Difference

Better experiments.
Stronger evidence.
Clearer decisions.

Any consultant can run a workshop. Here, you get 20 years of UX craft applied to experiment design, so the evidence you collect is actually strong enough to make a decision on.

Woman wearing headphones working on a computer with app designs displayed on screen in a bright room with plants and a sleeping dog nearby.

AI-native research methods

Prototypes isolate one variable. Interviews observe behavior, not opinions. You walk away knowing something real, not just what you already believed confirmed back at you.

Five colleagues collaborate around a wooden table with laptops and colorful sticky notes in a bright office filled with potted plants and whiteboards.

Your team reaches a decision

Structured sessions ensure that everyone's thoughts are visible, not just those of the most vocal person. Your team leaves with a decision they collectively support.

Smiling woman with curly hair and glasses sitting at a desk working on a laptop and large monitor in a modern office.

A framework that fits AI

Proven frameworks were rebuilt for model dependency, commoditization risk, and shifting user trust. You work with tools designed for the AI era.

Nice to meet you

Product designer and facilitator. 20 years running discovery sessions, design sprints, and strategic workshops across cybersecurity, pro audio, and telco.

UX background means I look at products through the user lens. Facilitation training means I run a room that surfaces what people actually think. Contributed to a $162M exit along the way.