MU20: Career Discovery for Students

MU20 prototype: The Entrepreneurs' Club page, with Collective, Career Launchpad and Skill Hub

the prototype that walked into a pre-seed funding meeting.

Pre-seed

the team raised it, Jan 2026 · I led product strategy

50+ → 3

features across 10 modules → 3 MVP modules (Discover · Do · Show)

student interviews that pivoted it. I ran six.

TL;DR

The Challenge

50+ features across ten modules, no conversations with students, and a pre-seed meeting coming. The real risk: building solutions before understanding the problem.

Team & Time

November 2025, a 4-week sprint. Me as product strategist, with a PM. We split the 12 student interviews, six each. The founders took the prototype into the room.

What I Did

Made the case for research, ran half the interviews, cut 50+ features to an MVP, and shaped six decisions, from indirect onboarding to the Opportunity Vault.

What Changed

Course-led became opportunity-led. Three modules survived: Discover, Do, Show. The team raised pre-seed in January 2026.

How do you want to read it?

01 – The Mess I Walked Into

50+ features. Zero conversations with students.

MU20 was my first company. So when they came back asking me to help define a product that could reshape how Gen Z students find career direction, I said yes without hesitation.

Try it · apply the filter

50+ features across ten modules.

OnboardingDiscoveryClubsOpportunitiesProfilesProgress trackingAI coachCompete layerCollege toolsAccount settings

“Does this help a student do something, or does it just tell them something?”

I made the case for user research. Not a long study. Just real conversations with real students before we committed to a direction. The team agreed. We talked to 12 high school students. A PM and I split the sessions, six each.

The team had momentum. What we did not have was a reason to build any of it.

02 – The Insight That Broke The Brief

We were building a course platform. They didn't want one.

We went into research thinking we were building a smarter way to deliver educational content. We came out knowing we were not building a course platform at all.

Student interviews · 12 students · tap a quote

What they said, and what it changed.

Not one student said they needed another course. They already had school for that. What they needed was closer to an action engine: a way to try things, not just learn about them. So we went from course-led to opportunity-led.

MU20 is an action engine, not a guidance engine.

03 – The Goals

Three questions every feature had to answer.

Not features. Not screens. Three questions that had to be true for the MVP to work.

Question 01

Does a student understand what to do on day one without anyone explaining it?

Question 02

Does the platform help them actually do something, not just browse?

Question 03

Do they come back on their own?

Everything we designed had to serve at least one of those three. If it did not, it did not ship.

04 – Key Design Decisions

Six decisions, defended twice.

Each decision had to work for two people: the student using it, and the investor judging it. Each one has a short log underneath with both defences. Try the demos.

DECISION 01

ONBOARDING

The psychometric onboarding.

Before a student touches any feature, the platform asks a set of indirect questions, designed with a child therapist's input.

Compare · the first question

Ask it straight, or ask it sideways.

Onboarding

What does the view from your dream room look like?

Describe itNext

Something more honest.

Indirect questions bypass that performance layer. The answers feed a profile that shapes every recommendation from then on.

Decision log

What it is

Indirect questions, not ‘what’s your dream job?’. The answers feed a profile that shapes every recommendation after.

For the student

Standard questions get rehearsed answers. Indirect ones bypass that performance layer and surface something more honest.

For the investor

Proof the team took the intake problem seriously and designed it with outside expertise. That intent on the first screen sets the tone.

DECISION 02

CORE LOOP

Micro Trials, a popular career format turned into a product.

Short, hands-on experiences that let students do the actual work of a career for two to four hours. The key choice: removing pass/fail entirely.

Try it · a Micro Trial

Do the work. Get a signal, not a grade.

Micro Trial · UX design

A usability task.

2 to 4 hours
No commitment
No formal enrollment

The skill-signal line is the example from the prototype. The trial tasks are the three examples from the brief.

Decision log

What it is

A trial of a career’s real work. The AI reads how you engage, not whether you got it right, and suggests what to try next.

For the student

‘Day in my life as a ___’ is one of the most popular kinds of career content. Students want to feel what a job is like, not be told. Without pass/fail, a test becomes a discovery.

For the investor

It fixes the biggest drop-off in career platforms: browse, feel overwhelmed, leave. And it creates the platform’s best signal: demonstrated tendency, not self-reported interest.

How I’d know

Activation: the share of new students who complete their first Micro Trial.

DECISION 03

FORK ON SCREEN ONE

The Explorer / Focus toggle.

At the start of onboarding, students choose their mode. Explorer opens a curiosity-driven feed. Focus fast-tracks students with a hypothesis to a domain deep dive.

Try it · the fork on screen one

One question. Two different first days.

Onboarding

How do you want to start?

Tap a mode on the phone.

Forcing exploration on a student who knows they want architecture is patronizing. Forcing commitment on a student with no idea is terrifying.

Decision log

For the student

One flow that treats everyone as undecided loses the students who already have direction. One question, two right-sized first experiences.

For the investor

Onboarding drop-off is where most EdTech products lose users for good. Getting the first experience right-sized is a signal of product maturity.

How I’d know

Direction confidence: a pre/post survey on ‘I know what to try next’, before and after onboarding.

DECISION 04

HOW WE TESTED

The Wizard of Oz, named.

In our Figma testing, the prototype was made to feel like an intelligent, personalized AI. Behind it, there was no model yet: just rules we set by hand. For the MVP, the team is building the real AI.

Peek · Figma testing

What the student saw: a recommendation that felt personal.

✦ Picked for youUX design · Micro TrialA usability task. 2 to 4 hours.

IF onboarding answers → visual, people-focusedTHEN show → UX design trialNo model. No training data. Just us.

The question wasn’t ‘can AI do this?’ It was ‘do students respond differently when it feels personal?’

Illustrative rule. This was only for testing the Figma prototype; the team is building the real AI for the MVP.

Decision log

Why fake it

Four weeks, no engineering runway. Faking the intelligence let us test the value of personalization before paying to build it.

For the student

We could watch how students responded to recommendations that felt personal, which was the real question.

For the investor

Pre-seed investors fund a thesis, not a tech stack. Naming the method out loud showed we knew the difference between validating value and building infrastructure.

DECISION 05

THE COMMERCIAL SPINE

The AI Pathway Builder.

A student picks a North Star. The system builds three paths toward it: one for cost, one for time, one for skill-building, each made of real, vetted opportunities in sequence.

Try it · the Pathway Builder

Declare a direction. The path builds itself toward it.

Your North Star“I want to work at a top design studio.”

Cheapest path · step 1 of 4

The bar above stays pinned across the whole platform, so a student always sees the next step.

Illustrative path. In the prototype, every step was a real, vetted opportunity from the Vault.

Decision log

For the student

Opportunity platforms show you everything and help you decide nothing. This inverts it: direction first, path second. Three variants, because a student in a smaller city on a tight budget needs a different path. That’s an equity choice as much as a UX one.

For the investor

This made investors lean forward. Committing to a path is a micro-contract with the platform: the base for retention, re-engagement and the B2B model.

DECISION 06

THE B2B SURFACE

The Opportunity Vault.

A searchable, filterable shelf of internships, fellowships, scholarships and competitions, curated by hand for launch. Each one carries a skill tag, a difficulty rating and a pathway-fit indicator.

Try it · the Opportunity Vault

One shelf. Two audiences.

InternshipDesign studio summer internshipVisual design✓ Fits your path

FellowshipYoung founders fellowshipEntrepreneurshipExplore

ScholarshipDesign merit scholarshipPortfolio✓ Fits your path

CompetitionNational pitch competitionPitchingExplore

CompetitionPoster design challengeVisual design✓ Fits your path

FellowshipPolicy & law research fellowshipResearchExplore

A curator’s shelf, not a job board: structured tags, soft cards, no urgency push.

Recreated for this case study. Listings are illustrative.

Decision log

For the student

Opportunity boards feel like job sites. The Vault feels like a curator’s shelf. Browsing should feel like exploring possibilities, not getting hired.

For the investor

It’s the revenue surface. Universities, fellowship programs and EdTech partners can sponsor slots inside student pathways. Designing the browse and the catalog as one object meant the economics scaled with usage, not headcount.

Show, in the prototype: The Vault, where a student’s wins and work live.

Six decisions. Each one served the student and the investor. Without compromising either.

05 – Built vs. Designed

A prioritization record, not a failure list.

What didn't make it wasn't a failure. It was cut on purpose to keep the core loop intact: Discover, Do, Show.

Your turn · make the call

Does it validate the core loop: Discover, Do, Show?

1 / 11

Psychometric onboarding, therapist-informed

The cuts followed one rule: does this exist to validate the core loop? If it didn't serve Discover, Do, or Show directly, it went to V2.

06 – The Outcome

In the room, with twelve teenagers between us and the check.

In January 2026, the founders walked into a pre-seed funding meeting with our high-fidelity prototype. They walked out with investment.

PRE-SEED FUNDED · JAN 2026

A prototype walked in. An investment walked out.

The design did specific work in that room. Psychometric onboarding reframed intake. Micro Trials demonstrated a new category. The Pathway Builder made the abstract concrete: here is what you do Monday morning.

The work I am most proud of is not any single screen. It is that the product we took to investors was fundamentally different from the product we started designing, because we stopped to ask twelve teenagers what they actually needed.

Outcomes

Measured

The team raised pre-seed in January 2026. Product metrics start when the MVP ships.

being measured — check backFirst Micro Trial completion, once the MVP is live

How I’d measure it

ActivationShare of new students who complete their first Micro Trial.
Willingness → paymentTrial-to-paid conversion once pricing is live.
Direction confidencePre/post survey on “I know what to try next”, before and after onboarding.

07 – What MU20 Could Become

The roadmap I'd build next.

The prototype helped get the round. The next three priorities turn it into a category.

NEXT · 01

A full recommendation engine.

Grow the MVP’s AI into a model that learns from psychometric signal, Micro Trial outcomes and pathway completion. The Figma tests showed students respond to personal; now make it cheap to deliver at scale.

NEXT · 02

A club community surface.

Students don’t just want pathways, they want peers walking the same path. A light community layer turns the platform from a tool into a place, and raises retention without raising acquisition cost.

NEXT · 03

A parent & counselor layer.

A low-touch dashboard for the adults already shaping a student’s decisions. Done right, it earns parent trust without taking the student’s agency away, and opens a second revenue surface.

08 – Reflection

What MU20 taught me about designing for funding.

Designing for an investor meeting is a different brief than designing for a daily user. The prototype has to hold up under a teenager's thumb and under an analyst's spreadsheet. Every decision had to be defensible from both sides of the table.

The biggest lesson was learning when to fake the infrastructure so you can prove the value. Calling Wizard of Oz by name kept the room on the right question: do students want this, before we ask, can we build it.

The goal was never to build a platform. It was to build a mirror.

50+ features. Zero conversations with students.

We were building a course platform. They didn't want one.

Three questions every feature had to answer.

Six decisions, defended twice.

The psychometric onboarding.

Micro Trials, a popular career format turned into a product.

The Explorer / Focus toggle.

The Wizard of Oz, named.

The AI Pathway Builder.

The Opportunity Vault.

A prioritization record, not a failure list.

In the room, with twelve teenagers between us and the check.

Measured

How I’d measure it

The roadmap I'd build next.

A full recommendation engine.

A club community surface.

A parent & counselor layer.

What MU20 taught me about designing for funding.

Related work