Field Note · Career Guidance · Part 2 of 3

From conversation
to prototype
to pilot

How we built Careers Ahoy! — and what each stage revealed about what career guidance is actually for

Location

Gurgaon, India — student sessions, parent interviews, flip cards on a table

The question

Can a tool help students discover their own path — without ever deciding it for them?

Read time

8 minutes

Series

Career Guidance · Part 2 of 3

The room · Gurgaon, 2024

"Don't make it fun. Make it realistic."

Five students. Printed flip cards spread across a table. A camera running. We'd spent weeks building to this — eight in-depth conversations with students and parents, a test of an AI recommendation engine, a 10-step content framework sketched and re-sketched. We thought we were testing whether the content was right.

The students told us something more important: they knew exactly what they needed from a career tool, and entertainment wasn't it. They were thinking seriously about their futures. The tool needed to match that seriousness. That sentence became the standard every subsequent decision was held against.

What the research showed

Career guidance fails when it treats a
discovery problem
as an information problem.

Before building anything, we tested the most obvious existing approach: an AI engine that assessed students and returned career recommendations. The output was coherent and broadly accurate. Two students. Immediate disengagement.

They didn't dispute the recommendations. They felt no connection to them — no ownership, no agency. One nodded, said she'd think about it, and moved on. Technically responsive. Emotionally absent.

The problem wasn't the AI. It was the model: the assumption that students needed a better answer. What they wanted was a structured way to arrive at their own.

The shift that drove every design decision

Better answers, faster

→

Frameworks for self-discovery

"Here's what fits you"

→

"You decide what fits"

One session, one output

→

A journey through one career at a time

Student evaluated

→

Student in control, always

Reassuring, promotional tone

→

Honest about tradeoffs — always

The lab process

Seven stages from a first conversation
to a pilot-ready product

01 · Listen

Eight conversations before building anything

Career anxiety was active and unnamed. Students were paralysed not by too few options — but by too many with no personal filter.

02 · Test the obvious

The AI engine that worked — and didn't

Coherent output. Immediate disengagement. The recommendations were plausible. The connection wasn't there. Students received an answer. They wanted to find one.

03 · Sketch

A 10-step journey through one career

Not a comparison tool. A portrait. Honest enough that a student could form a genuine opinion — including that this career is wrong for them.

04 · Prototype

Flip cards, five students, one camera

Engagement shifted immediately. Students processed out loud, navigating rather than receiving. Sharp feedback: more depth, more realism. "Make it realistic."

05 · Framework

Hundreds of careers. Eight archetypes.

Career domains required prior knowledge to navigate. Archetypes — orientations, not job lists — gave students an entry point that felt instinctively right.

06 · Build

Three pathways, fully built for pilot

SE, UX/Design, Clinical Alt Medicine. One familiar, one creative-technical, one completely outside the usual frame. The hands-on simulations emerged here.

07 · Ship

Vibe-coded. Mobile-first. One pivot changed everything.

Built with AI tools — no prior technical knowledge. Web-to-mobile changed how information was chunked, navigated, and felt entirely.

Student session · Gurgaon, 2024

Watch the shift from receiving an answer to finding one

The difference in engagement between the AI output test and the flip card session was immediate and visible. Students slowed down. They asked questions of themselves, not us — processing out loud, navigating at their own pace. The structure of "here is the reality — now what do you think?" created something the AI never did: a genuine sense of agency over the decision.

This clip features students. We have blurred faces to protect their identity.

The entry point

Eight archetypes.
One question that opens everything.

Career domains are useful — but they require prior knowledge to navigate. Archetypes ask something simpler: what draws you toward certain kinds of work? The answer is instinctive. The world opens from there.

⚒

Builders

You like to create things from scratch

◎

Designers

You like to bring improvements through better design

♡

Carers & Supporters

You like to support, care for, and serve others

◈

Protectors

You like to protect, uphold, and keep things safe

⊞

Managers & Improvers

You like to organise and make systems better

◐

Analysts

You like to investigate, research, and find patterns

⊕

Sustainers

You like to protect and preserve what matters

☆

Showcasers

You like to perform, create, and be seen

Inside each pathway

Ten stops. One complete portrait of a career.

Each pathway follows the same structure — but the student's own reflections at every checkpoint make it personal. The journey is designed. The destination is theirs.

Say Hello to the Pathway

First impressions · What is this, really?

A Day in the Life

What a real working day looks like

What Is Work-Life Like?

Honest about the tradeoffs

What You Need to Enter

Skills · Pathways · Real requirements

Pause & Reflect

Midpoint · How is this landing?

Specialisations

The range of what this career can become

AI's Impact

What's changing · What stays human

Meet People & Get Inspired

Real people doing this work

Hands-On Activity

2 simulations · Do the thinking the career requires

Final Reflection

Bag it, maybe, or move on

The digital product

Careers Ahoy! —
built for the phone they already have

The web-to-mobile pivot changed how everything was designed. Information that worked as a paragraph on a desktop became a card, a swipe, a single question at a time. The phone is where these students live — designing for it meant designing for how they actually think and move through content.

The product begins with a short onboarding, moves to archetype selection, and then into the chosen pathway. Data capture runs throughout — tracking where students slow down, where they exit, how they rate each stop. That data is the foundation of Part 3.

Careers Ahoy! home screen — archetype selection

Home · Choose your archetype

Careers Ahoy! — PizzaBot 3000 hands-on simulation

Stop 09 · Hands-on simulation

Careers Ahoy! — Pause and Reflect checkpoint

Stop 05 · Pause & Reflect

The hardest design problem

Staying in discovery mode —
not slipping into
prescription

The temptation to become prescriptive appeared at every stage of the build. Write a skills section and it drifts toward "here's what you need to develop." Design an activity and it starts to feel like a competency test with a verdict underneath.

PizzaBot 3000 — the Software Engineering simulation — could have scored students and nudged them toward a conclusion. We made it a reflection prompt instead: did you enjoy the step-by-step logic? Did the bugs feel like a fun challenge or a source of real frustration? No right answer. The point is that the student finds out something true about themselves.

The broadest version of this principle runs through everything. The job isn't to tell a student what career suits them. It's to create the conditions in which they can work that out — and trust their own answer when they get there.

"There was no one to guide me. I was always scared I would regret my decision."

Student · Early research conversation · Gurgaon

"Exploring by ourselves is definitely better."

Student · After the prototype session

Early responses

What people said when they used it

"Love how interactive the app is. Very playful."

Sanat Sarin
Recent design graduate

"The content has been well thought through. The simulations are fabulous."

Tushar Dadlani
Tech entrepreneur, US

"It would be a blessing for students from government schools where career guidance is unheard of."

Chapter Convenor
Shalini Fellowship

From conversationto prototypeto pilot

Career guidance fails when it treats adiscovery problemas an information problem.

Seven stages from a first conversationto a pilot-ready product

Eight archetypes.One question that opens everything.