Rhym
ROLE :
Product Design, Prototype
Development
TIMELINE :
5 weeks
TOOLS :
React, Node.js, OpenAI, Replit, Miro
Project Overview
Context :
This project was developed during a 5-week business design sprint at Central Saint Martins by a team of 4, with the challenge of building an end-to-end product around the theme of employee wellbeing.
The problem:
Freelancers have autonomy over when they work, yet 76% take less time off than regular employees. Without external structure, they are often left to create their own systems for managing work. While productivity tools help organise tasks, they rarely account for how a person’s energy fluctuates throughout the day. As a result, freelancers often manage their time but struggle to understand their energy.
What we did:
We created Rhym, an AI-powered desktop companion that helps freelancers align their work with natural energy rhythms through ancient wellness practices rather than rigid time-based productivity systems.
RESEARCH :
Even though 73% of Gen Z and independent workers are increasingly seeking body awareness and sustainable productivity systems, the current market landscape fails them because it is siloed. The market currently operates in three separate ecosystems:
This creates a strange experience where:
• Productivity apps treat you like machines
• Health tracking generates data — but don’t connect it to work
• Wellness practices are separated from everyday life, often reserved for retreats or occasional activities
OPPORTUNITY :
Rhym sits at the intersection of these three industries, where productivity, health awareness, and embodied rhythm (wellness) can exist as a single layer within everyday work.
CONCEPT :
Rhym is a desktop companion that helps digital freelancers reduce the mental load of planning work while gaining insight into their natural productivity rhythms.
Instead of organising work purely around time, Rhym introduces the idea of body intelligence — recognising internal energy signals to better understand how and when we work best by remembering your work patterns.
We built on the idea of “body intelligence” - defined as recognising internal energy signals to enable users to understand their productivity rhythms. we do so by reflections and nudges based on ancient wellness practices.
Research shows that wellness practices are effective at managing daily energy slumps and long-term burnout and can see a productivity increase of up to 20% (World Health Organisation).
Try the live project, works best on desktop :)
BUILDING THE PROTOTYPE :
The fun part (for me)
As I have always been an ideas girl, I started experimenting with vibe coding because it gave me the power to ship, test, and iterate quickly on ideas I would otherwise only speculate about.
Rhym was built using AI-assisted development on Replit, with:
• a React frontend
• a Node.js backend
• AI capabilities powered by OpenAI
PRODUCT DEVELOPMENT CYCLES :
Product Iteration #1 - getting the simpest form of the idea out
I built a browser extension with simple task and energy tracking, all user-input based to understand how workflows realistically function.
design decisions
When people feel overwhelmed, they don’t necessarily stop working — they change how deeply they engage with tasks.
we often underestimate or overestimate task duration, this feature helps provide insight into that gap
The extension gave users agency to choose their energy levels and prioritised tasks accordingly
Product Iteration #2 - Leveraging AI for insights
We evolved the product into a desktop app that simplified the experience into three main interactions:
• Rambling about what's on your mind - from productive tasks to mundane things like walking your dog or calling your mom.
• Rhym reading your energy level and generating a task list ready to act on.
design decision
reflection and nudges which bring in wellness in a grounded practical way.
• Integrating with AI - in order to provide insights that bring wellness into your workflow and create a deeper understanding of your energy patterns.
Product Iteration #3 - Giving depth to the product idea
Our understanding of the problem evolved and we began exploring ways to incorporate knowledge from ancient wellness practices such as:
• Ayurveda
• breathwork
• Chinese medicine
alongside modern research on circadian rhythms.
The goal was to allow users to dive deeper into understanding their own energy systems and navigate work alongside different health or body-related challenges.
We introduced:
-
Rhymedies : a knowledge repository exploring these themes.
-
Rhymeets : a community feature designed to encourage conversations about work and wellbeing.
BUSINESS THINKING :
We explored a tiered subscription model after analysing the pricing strategies of competing products.
The long-term vision includes integrating third-party health tracking devices into Rhym to generate deeper and more personalised insights.
CHALLENGES & LEARNINGS:
Prompt design
Achieving useful and consistent AI responses required experimenting with prompt structures and being very intentional with wording.
Token efficiency
Working with AI highlighted the importance of managing token usage and keeping prompts concise to control API costs and speed up testing.
Data privacy considerations
Because user inputs are processed through an AI API like OpenAI, I had to carefully think about minimising the amount of sensitive data sent externally.
Scope control
Working within a five-week sprint required focusing on an MVP rather than attempting to build every possible feature.
Personal reflection
Working on the logic and reasoning of the product first, rather than the visual look of things, felt refreshing.
It gave me a newfound respect for engineers and allowed me to use AI as an extension of my own cognition.
TIMELINE & PROCESS :
Week 1 : Started with qualitative primary research (16+ participants), mapping behavioural insights and opportunity gaps.
Week 2 : Analysed the competitive landscape and vibe-coded a high-fidelity prototype.
Week 3–5 : We extensively used our own product and ran three rounds of product testing, which helped simplify and logically sequence features.
The final idea was shaped by three key discoveries in the process:
01.
Industry early signals
The wellness-tech market continues to grow rapidly by 2030, particularly among digital-native professionals (Global Wellness Institute)
Informed field research assumptions
02.
Field research insights
"Even if my energy slumps, my work doesn't stop. sometimes i dont have the brain capacity to think and delegate tasks."
cognitive overload simplified with features
03.
Design ethic
How can we retain users’ agency while a machine makes inferences about their productivity?
shaped the design principles in user interaction
next up is a LEGOSET in 2040