I turn messy problems
into delightful experiences.

The Spark

A few days before building this project, I experienced a setback that felt much bigger than it really was. In that moment, I completely overlooked everything I had achieved and judged myself solely based on that single experience. Thankfully, conversations with my friends and family reminded me to zoom out and look at the bigger picture. They encouraged me to remember every milestone, not just the moments that went wrong.

That experience made me wonder: what if we had a dedicated space to collect our achievements, so we could revisit them whenever we needed perspective? That question became the foundation for Achievement Diary.

The Problem

When people experience setbacks, they often struggle to recall evidence of their own growth. Negative experiences become more mentally available than positive ones, making it easy to feel like no progress has been made. While many journaling apps focus on daily writing, gratitude or mood tracking, very few are designed around building an evolving archive of achievements that users can revisit during moments of low confidence.

How might we help people make their progress visible, so that one difficult day doesn't erase months or years of growth?

Design Principles

Reflection builds self-efficacy. Albert Bandura's Self-Efficacy Theory suggests that recalling previous successes strengthens confidence in tackling future challenges. Instead of relying on memory alone, Achievement Diary provides a tangible record of personal accomplishments that users can revisit whenever they need reassurance.

Celebrate progress, not perfection. Behaviour change research shows that recognising progress increases motivation. Rather than encouraging users to achieve more, the product encourages them to notice what they've already accomplished.

Remove friction. Reflection is only sustainable if it feels effortless. The experience was intentionally designed to minimise cognitive load through simple authentication, fast achievement logging, clean distraction-free layouts, and a chronological view that makes progress easy to revisit.

From Idea to Working Product

I first mapped out the core user journey: create an account, record achievements, build a growing personal archive, and revisit previous milestones during moments of self-doubt. Keeping the scope intentionally focused allowed me to validate the core idea before introducing additional features.

Using Google Stitch, I generated the initial interface concepts, experimenting with layouts that prioritised readability and reflection. Rather than accepting AI outputs as final, I iterated on the structure, hierarchy and interaction patterns to create a calmer, more focused experience. I then used Lovable to translate the product vision into a functioning web application, implementing authentication, achievement creation, navigation and overall application flow through iterative prompting, testing and refinement. Supabase handled user authentication and securely stored achievement entries, letting each user build a private, persistent diary over time.

Outcome

The final product lets users create an account, securely save achievements, build a growing timeline of personal milestones, and revisit their progress whenever they need perspective or encouragement. More importantly, the project shows how behavioural psychology, UX thinking and AI-assisted development can work together to turn an idea into a functioning product within a matter of days.

What I Learned

This project reinforced that AI is accelerating how products are built, but not why they should exist. While tools like Google Stitch and Lovable dramatically reduced development time, the most valuable work still involved defining the problem, grounding the solution in behavioural principles, making design trade-offs, and continuously refining the experience.

The Problem

Students are constantly told to budget, but budgeting apps feel like homework: spreadsheets, guilt, red warning screens. Most students aren't broke because they can't track spending — they're overwhelmed and emotionally disconnected from money, so they avoid it entirely and never build a savings habit. The problem isn't information. It's feeling.

Research

Before designing anything, I surveyed 16 Gen Z students and young adults on the emotional reality of saving.

10 of 16 already save every month or most months — people are doing the behaviour and hating it. Visual growth of savings was the highest-rated feature in the appeal grid. One respondent put it best: "If AI could grab our habits of expenditure but not track transactions, it could help while having our own privacy of expenditure. You don't wanna be in a phase where you hide your expenses from AI." That sentence became a design principle: intelligence without surveillance.

Theoretical Grounding

Fogg's Behaviour Model (B=MAP): a behaviour forms when motivation, ability, and a prompt align. Most saving apps destroy ability through complex setup and ignore motivation through guilt. Bloom inverts this — the daily seed is the prompt, the AI companion sustains motivation, and starting at 50p removes the ability barrier entirely.

Klontz Money Scripts: people's relationship with money is driven by belief patterns, not lack of knowledge. This grounded Bloom's personality onboarding — understanding who you are with money, with compassion, before asking you to save anything.

The Solution

Bloom is a gentle savings companion that turns saving into a daily ritual. A short personality quiz reveals your financial personality type and shapes how Bloom speaks to you. You set a goal and save in tiny daily steps starting at just 50p, watching a flower grow in your pot as you go. A friendly AI companion, by voice or text, surfaces gentle reflections and spots patterns in your relationship with money over time. Friends can join the journey too, so you're never relying on willpower alone. Reach Day 15 and your pot blooms a real reward.

Key Design Decisions

What Was Built

Personality-based onboarding, a growing savings pot with six visual stages, a daily seed-planting ritual with streak tracking, an AI companion with voice and text input, friend-based motivation, a Day 15 reward milestone, and a Supabase backend with real user accounts.

Product thinking and strategy happened in Claude, working through research, theoretical grounding, feature prioritisation, and design decisions before a single screen was built. Having a thinking partner at that stage meant arriving at Lovable with clarity rather than confusion: a defined problem, a defensible solution, and a clear scope for what v1 needed to be.

Bloom was built and submitted as part of SheBuilds 2026, a hackathon run in partnership with Lovable. I was selected to participate from over 3,000 applicants, with only 52 places available globally. Being selected was validation that the problem and the idea were worth taking seriously.

Reflection

The most important design decision was a restraint: not building a budget tracker. Every time the product drifted toward tracking, it drifted away from what made it worth building. Saving should feel like growing something, not accounting for something.