It is hard to become what you cannot see.
Elevra matches young women with STEM role models who share their specific lived experience — culture, location, disability, identity — instead of the same ten famous names on every poster.
Aaliyah Rahman
AI Engineer at CSIRO — failed Year 11 maths, now builds language models.
"You don't have to be the smartest in the room. You just have to be the one who keeps showing up."
A business problem, walked back to a human one.
Enterprise Design Thinking argues a problem well-stated is half-solved. Here's how a stated pipeline gap becomes the problem Elevra actually solves.
- Girls leave the STEM pipeline at a disproportionate rate — worse for intersectional groups.
- They don't see a version of themselves persisting in a STEM career.
- The role models they're shown don't reflect their specific background.
- Programs default to the same small pool of well-known women, because manual matching doesn't scale.
- No tool matches role models to students by lived-experience overlap, not gender alone.
- So what: build the matching tool — and treat "no match found" as signal, not silence.
Two doors in, one honest engine behind them.
Swipe if you're just looking. Answer five questions if you want precision. Either way you land on the same transparent matching layer.
Two ways in
Swipe through stories with zero commitment, or run a 60-second identity quiz for precision matches — no one is forced to disclose before they're ready.
Transparent matching
A weighted tag-overlap score, not a black-box model. Rare identity combinations are weighted higher, and the "why you matched" reasoning is shown in plain language.
Her Path timeline
Every story includes the barrier she faced at the user's exact age, the turning point, and a visual milestone timeline to her current role.
Missing Story alert
A zero-match search isn't a dead end — it's logged as a recruitment brief and surfaced to program coordinators.
Smart Apply
Scholarships and programs filtered to your profile, with a LinkedIn import, resume upload, or a guided no-resume conversation for students who have neither yet.
Coordinator dashboard
Organisers see which identity combinations have demand but no stories available — the gap becomes a to-do list, not a mystery.
The prototype, running right here.
22 working screens, no backend required
Built as a real clickable HTML/CSS/JS product for the hackathon demo, not a static mockup — every screen below is interactive inside the frame.
- Splash → pick your path → swipe or match quiz
- Your matches → story detail → her path timeline
- Missing Story alert for zero-match identity combinations
- Take-action hub → program detail → Smart Apply
- Saved library, profile, and coordinator dashboard
Warm, not clinical — on purpose.
Terracotta and sage instead of the standard blue/purple SaaS default. This is a product about seeing yourself in someone else, not a dashboard.
One sprint, run as a real design-thinking loop.
Built in ~24 hours by a 4-person team — an Agile sprint, not a phased plan, since requirements were being discovered in real time.
Conceptualise
Reframed the brief into a Hills statement (Who/What/Wow).
Iterate
Two concepts explored in parallel — swipe browser vs. identity-match engine.
Converge
Merged into one product: two doors, one shared story layer.
Build
Built as a real clickable prototype, not static mockups.
Pitch
Scripted a 4-speaker, 3-minute pitch around one persona's journey.
The full write-up.
Problem & Research
5 Whys, Hills, personas, HEART metrics, risks.
DesignDesign & UX
Principles, user stories, journey map, IA, design tokens.
ProcessRoadmap & Process
How it was built, prioritised backlog, risk register.
HistoryProduct History
How two early concepts merged into Elevra.
PitchPitch Script
The 3-minute, 4-speaker hackathon pitch.
SubmissionDevpost
The official hackathon submission page.