Talk with an AI coach that listens, asks follow-ups, and points out what you haven't noticed — then walk away with an analysis you can keep.
Pick the question you most want to clarify right now. You do not need a ready answer — just follow the guidance once you are in.
Talk through your interests, what matters to you, and what you have actually done, then leave with a few career directions worth testing first.
Use delegation, feedback, and team conflict scenarios to see your first instinct — and whether you actually want this path.
See how you actually work with AI: how clearly you brief it, whether you check what it produces, and which calls you keep for yourself.
Inside the assessment you meet an AI coach that works like a mentor sitting beside you: it organizes the clues you give, points out what evidence is still missing, flags the things you may not have noticed — and asks only one key question at a time.
A rough opening is enough — just say whatever comes to mind. The coach sorts out what is known and what is still missing before deciding what to ask.
Some signals distort easily when described directly, so Compass uses trade-off scenarios instead: two conditions pull against each other, and your ranking records your priorities. These scenarios are AI-generated too, and there is no correct answer.
At the end of every stage, the coach writes down the signals, the evidence you provided, and the points that still need cross-checking. All of it feeds the next stage, so the conversation keeps getting closer to your situation.
CHECK IT YOURSELF
Most tests hand you a result and ask you to take it on faith. Ask one why it says that about you and it has no answer — the verdict was computed from a pile of multiple-choice answers, with a layer of math you never get to see sitting between the questions and the conclusion.
Here, every conclusion opens up. Click it and you see which words you said, at which point in the assessment. You may disagree with the inference — but you can always see where it was drawn from.
The report stops at hypotheses on purpose — the decision stays yours. What it gives you is evidence, observations, and actions small enough to actually try.
Open prompts accept vague replies. The coach restates what it heard, keeps multiple hypotheses open when evidence is thin, and holds off on conclusions.
Your answers and trade-offs from different stages are laid side by side. Recurring patterns surface — including the one you never said out loud but kept repeating.
Every conclusion comes with a low-risk next step: perhaps a conversation, a small sample, or a tiny experiment. Trying it is how you find out whether the observation holds.
Each topic has its own stages and scenarios, but the rhythm is the same: the AI generates a scenario, you respond, the AI closes the stage into signals — accumulating stage by stage until everything is cross-checked into a report.
Choose the question closest to your situation. No preparation needed — answer on instinct.
The interaction alternates: situational choices, rankings, and short dialogues. Even the options are generated live from your earlier responses — no right answer, just a record of your trade-offs.
The AI summarizes the evidence and open questions from this stage and passes them forward, so later scenarios fit you better.
The report is written as hypotheses, each spelled out: the evidence behind it, what still needs checking, and a small task to try first.
Every assessment topic has a fixed measurement blueprint that defines which signals to observe. The AI turns that blueprint into this session's scenarios, interactions, and report — which is why no two people go through the same session.
The problems Compass handles share one trait: the information is too messy to make a decision yet.
Something is on your mind but hard to put into words — a stale job, an itch you keep second-guessing, a role you play with less confidence than you show. That vague description is exactly where an assessment starts.
Four letters or a single score never quite match the person you know yourself to be. A conversational assessment treats your own words as evidence, so the result resembles you — the material is you.
Thinking about yourself tends to go in circles. The AI coach has no agenda; it works only with the clues you give and lays out the patterns that keep repeating.
Compass is a tool that uses AI to help people talk through unclear directions. You answer in your own words, it asks a few follow-up questions, and it organizes the thoughts, hesitation, and details you may have missed.
Typical tests draw from a fixed question bank and ask people to choose from fixed options. Compass generates its questions live from what people actually say, gives them room to explain their reasons, and follows up from there.
Compass is useful for people thinking through career direction, a role change, a first job, or a field of study. Founder fit and management style topics are still in design and will use the same conversational format once open.
AI models do embellish, so a second check runs before any report is written.
A separate AI receives the conversation transcript and every claim the first AI made. It answers one question: is this claim grounded in something the person actually said? Claims that fail are dropped.
Every supporting point in the report opens up to show the person's own words and the stage where they said them. A direction that cannot be grounded is left out, and the report says so plainly.
Compass organizes the directions and next steps it can see. The choice still depends on real-life testing, personal context, timing, and what the person wants to carry forward.
START WITH WHAT YOU KNOW
Pick the topic closest to where you are. Rough answers are fine — the coach fills in the gaps with you. And whatever observations you take away, pick one and verify it in the real world.
