Melbourne-based · Open to full-time roles and selected projects

Chief of Staff with a background in product and engineering.

Over eight years at a Melbourne product studio, my work grew from product discovery and delivery into customer relationships and helping run the business.

Startmate First Believers

01Where I fit

Where I can add value

My work has crossed these four areas, often at the same time. The roles that suit me best tend to combine a few of them.

Chief of Staff / GM

Working alongside founders and leadership teams to turn priorities into action, keep people aligned and make sure important things get followed through.

Product & Delivery Lead

Taking digital products from early discovery through design, development and launch, bringing customers, designers and engineers together around what needs to be built.

Client Partnerships & Delivery

Building long-term customer relationships, navigating competing priorities and difficult conversations, and following through on commitments from discovery to delivery and beyond.

Teams, Systems & Operations

Helping teams do good work through clear ownership, sensible planning, coaching and just enough structure to keep things moving.

02The receipts

What I'm proud of

Eight years at Launchpad App Development, progressing from product management into customer, delivery and company leadership. My contribution was building and leading delivery teams, customer relationships and internal systems that supported the studio's growth.

5→30

Launchpad's growth from five to thirty people

~$2M

Annual work value across the strategic accounts I led

Concurrent strategic accounts at peak

5★ Clutch reviews from customers whose accounts I led

50+

Product engagements I worked across

Certifications I led for the business, end to end

03Right now

Currently building

Products and ideas I'm developing, from an active build to early interactive concepts.

Velo Passport

A digital passport for the places cycling takes you. Collect stamps, revisit the places you've explored and build a record of your cycling journeys. A personal product born from a year riding around Europe, now moving through design, prototyping and AI-assisted development.

View case study →

ConceptInteractive prototype

Moss

Nervous System Companion

A small companion for burnt-out nervous systems.

View case study →

ConceptInteractive prototype

Keeplight

Energy Companion

A companion for life with energy-limiting illness.

View case study →

04Work with me

Work with me

I also take on a small number of project and fractional engagements with founders and product teams.

Zero-to-One Product Sprint

From idea to something you can see, test and confidently build.

Product & Delivery Reset

Find what's not working, fix what matters and get moving again.

Fractional Product & Delivery Partner

Senior product and delivery support, without another permanent hire.

What I'm looking for

Open to the right full-time opportunity.

I'm looking for the right next opportunity across product, client delivery or business operations.

I value good people, real ownership and interesting problems to work through together. The exact title matters less to me than the team, scope and opportunity to contribute.

Melbourne-based and open to hybrid or remote roles where the fit is right.

Get in touch →

Selected work

Products, systems, and proof of work.

A selection of products I've built, company-wide programs I've led, and client products I've delivered over the past eight years.

01Product builds

Things I've built myself

From personal projects to tools that solve problems I've encountered along the way. Some are prototypes, some are live, and all have taught me something about building products.

Flagship build ● Active · pre-launch

Velo Passport

A digital passport for the places cycling takes you.

Inspired by a year riding around Europe, Velo Passport is designed to turn everyday rides into a record of the places you've explored, collecting stamps along the way.

I'm building it from the ground up, working through customer discovery, product decisions, UX, visual design and AI-assisted development.

View case study → velopassport.co ↗

Design prototype · sample data

Product concepts · prototyped, not yet tested

ConceptInteractive prototype

Moss Nervous System Companion

A small companion for burnt-out nervous systems. Say how you're doing and get one thing to try that fits what you have left. Seasons, not levels, and nothing to keep up with.

View case study →

ConceptInteractive prototype

Keeplight Energy Companion

A companion for life with energy-limiting illness. It remembers who you are beyond illness and expects nothing in return. No tracking, scores or goals.

View case study →

Supporting builds and experiments

Build ● Active AI workflow

DevOS - AI-assisted build workflow

An evolving workflow I've put together to build products with AI, drawing on the discovery, delivery and quality practices I used at Launchpad.

It takes a project through discovery, specification, design, development, testing and release, with different AI agents handling defined tasks and review points where I make the decisions.

Each build helps me refine the process for the next one.

Idea → release · ✓ = manual review point

PHASE 0 Orchestration orchestrator

PHASE 1 Intake clarifier

PHASE 2 Discovery & Spec product-owner ✓

PHASE 3 Architecture architect ✓

PHASE 4 Design System design-system ✓

PHASE 5 Branding branding ✓

PHASE 6 Assets assets

PHASE 7 Features feature-logic ✓

PHASE 8 UI ui

PHASE 9 Landing landing-page ✓

PHASE 10 User-Sim user-sim

PHASE 11 QA & Security qa-security ✓

PHASE 12 Growth growth ✓

PHASE 13 Release release ✓

PHASE 14 Retrospective retro ✓

↻ Retrospective - update the workflow after each build · Documentation runs across every phase

15 phases 16 specialist agents Manual review points Versioned workflow

WEIGHTED SCORE

Build● Live demo

Cohort Shortlist Ranker

A small decision-support tool I built for my own Startmate First Believers review process. It combines weighted scoring, manual ranking and written reasoning to help compare applicants without letting a numerical score make the final decision.

Demo runs on fictional sample companies.

Try the live tool ↗

78 ROUTES MAPPED 3,658 KM ACROSS FRANCE, BELGIUM & THE NETHERLANDS

Build● LiveStrava API

Cycling Dashboard - The Europe Year

A decade of cycling data, and a way to revisit where it's taken me. I built an interactive dashboard using my Strava history, bringing together ride maps, European trips and comparisons of my cycling before and during our year overseas.

Built as a self-contained HTML dashboard using the Strava API, Leaflet, SVG and vanilla JavaScript.

View dashboard ↗

02Company-wide programs

Company-wide programs and internal systems

Work I led across Launchpad as the team grew from five to thirty people, covering internal operations, quality and security, customer delivery and the eventual company wind-down.

Led

Customer Success & Client Relationships

Built the customer rhythms, feedback loops and escalation practices that helped support long-term partnerships across roughly 15 concurrent strategic accounts. Customers whose accounts I led contributed 26 verified five-star Clutch reviews.

Led · Scale-up

Company Operations & Internal Systems

Built and improved the systems behind how the studio worked, from planning and resourcing to onboarding, coaching, documentation and team communication. The aim was to give people clarity and ownership as the business grew, without creating unnecessary process.

Led

Quality, Security & Culture

Led the studio's B Corp, Great Place to Work, ISO 9001 and ISO 27001 programs, working across teams to turn commitments around quality, security and culture into practical ways of operating.

Transition · High-stakes

Company Wind-Down & Client Continuity

Stepped into operational leadership through a difficult period for the business, coordinating restructuring, client transitions, team communication and project handovers. The focus was on being transparent, maintaining continuity where possible and looking after the people affected.

03Client delivery

Products I helped bring to life

A selection of digital products I worked on across more than 50 client engagements at Launchpad. My involvement ranged from hands-on discovery and product management through to leading delivery teams and customer relationships. These were multidisciplinary efforts, and I've called out my role in each.

Healthcare · Paediatrics

Ollie Pediatrics

Patient management & telehealth platform

My role · Product management · Patient and provider workflows · Client partnership

A paediatric healthcare platform covering appointment booking, telehealth, symptom and medication tracking, secure communication and health records. The product brought multiple parent and provider workflows into one system, with the additional constraints that come with healthcare data, privacy and trust.

Regulated healthcare data · Trust & safety critical olliepediatrics.com ↗

Healthcare · Surgical care

Mend

Pre- & post-operative patient recovery platform (ERAS)

My role · Product delivery · Care pathway workflows · Client partnership

A platform for improving pre- and post-operative care, guiding patients through preparation before surgery and structured recovery afterwards. The work sat across clinical workflows, patient behaviour change and the compliance requirements that come with care delivered outside the hospital.

Clinical workflows · Behaviour change · Compliance mendapp.au ↗

Operations dashboard · client details redacted

Asset monitoring dashboard with alarm, fault and emerging status cards, telemetry charts and a scheduled maintenance panel

Industrial · Fire safety

Predictive Asset Platform

Predictive asset management & monitoring platform

My role · Technical delivery · UAT · Stakeholder coordination · Rollout

A connected asset-management platform for industrial fire-suppression equipment, spanning technician and configuration apps, web and tablet interfaces, cloud monitoring and predictive fault detection. The delivery involved connected hardware over Wi-Fi, Bluetooth and MQTT, commissioning and service reporting, enterprise UAT and development of the ML fault-prediction capability.

Wi-Fi · BLE · MQTTEnterprise UATML fault predictionWeb · tablet · technician & config apps

Safety-critical systems · ML-driven fault prediction · Multi-role ecosystem

Dial Before You Dig app showing a lodged job with its mapped dig area

Public infrastructure

BYDA / Dial Before You Dig

National geospatial request & compliance platform

My role · Product delivery · Redesign · Integrations · Launch

Led product delivery on an earlier redesign of Dial Before You Dig's national mobile application, including its first Android release and integration with Esri geospatial services.

1,300 launch-day downloads100,000+ users2M+ requests in the first 12 months

Public infrastructure · Scale · Geospatial data byda.com.au ↗

Children's health & fitness

BrickFit

Gamified physical activity program

My role · Product strategy · Delivery planning

A children's fitness product combining physical activity challenges with LEGO building rewards, developed with LEGO Certified Professional Ryan "The Brickman" McNaught. The product included activity challenges, progress tracking, badges and builds, friends and groups, family accounts and an administration platform.

Engagement design · Child safety · Motivation brickfit.com.au ↗

Government · Preventative health

Preventative Health Platform

Behaviour-change & health tracking platform

My role · Product delivery · Customer engagement

A behaviour-change platform designed to help people make and maintain healthier lifestyle choices while giving healthcare providers a clearer view of patient progress. It included personalised health plans, progress dashboards, family and community challenges and communication between users and their care teams.

Longitudinal engagement · Data-driven care

Health · Aged care

Memory Machine

Assisted living & memory support system

My role · Client partnership · Product delivery · Ongoing support

Assistive technology for people living with memory loss and the family members and carers supporting them. The platform combined reminders and events, item tracking, video calling and shared carer workflows across tablet, mobile and web experiences.

Accessibility · Care network coordination memorymachine.com.au ↗

04Tools

Tools I've worked with

Tools I've used across product delivery, customer work, company operations and the projects on this site. The AI tools are newer, mostly from my own recent builds. Jira, Basecamp, Notion and the customer systems were part of my day-to-day work at Launchpad.

AI & automation

Claude CodeCoWorkChatGPTZapier

Product & delivery

JiraBasecampSlackMiroCanvaGitHubFigmaGCP

Knowledge & customer ops

NotionLoomGoogle WorkspaceSalesforceHubSpot

About

The through-line.

I started in mechanical engineering and found my way into digital products. Over eight years at Launchpad, I went from shaping early-stage products and working with designers and engineers to leading customer relationships and helping run the business.

I enjoyed working across disciplines, being close to the people doing the work and getting involved wherever I could be useful.

Now

Melbourne

Back in Melbourne after a year in Paris, building Velo Passport, reconnecting with the local product and startup community and exploring what comes next.

2026

Startmate First Believers

Reviewing early-stage startup applications alongside founders, operators and investors. A chance to look at product ideas, customer evidence and business models from the investment side of the table.

2025 - 2026

A year in Paris

A year based in Paris, exploring Europe, riding bikes and enjoying a change of pace. Somewhere along the way, an idea inspired by all that cycling became Velo Passport, giving me a reason to get hands-on with product design and AI-assisted development.

2017 - 2025

Launchpad App Development

Lead Product Manager → Head of Product Delivery & Customer Success → Chief of Staff / GM

Across eight years, I worked on more than 50 digital product engagements, built long-standing customer relationships and helped grow the business. The roles changed, but much of my work centred on connecting people, solving problems and keeping things moving.

2016 - 2017

Business & Data Analyst, Futures Sport & Entertainment

Worked in business and data analysis, turning sporting and media datasets into commercial insights and sponsorship recommendations.

2013 - 2016

Graduate Mechanical Engineer, Aurecon

Started in mechanical design, project engineering and procurement on Santos projects. Alongside that work, I completed a Graduate Diploma in Innovation & Entrepreneurship (Technology & Business), adding a commercial and business foundation to my engineering background.

How I work

A sounding board

I like helping founders and teams think decisions through. Sometimes that means being a sounding board, having a difficult customer conversation, asking the questions nobody's asked yet or helping a team find a way forward.

People first, honest conversations

I put people first, but that doesn't mean avoiding difficult decisions or conversations. I believe in listening carefully, being honest when something isn't working and following through, particularly when things get difficult.

Just enough structure

I'm a fan of just enough structure to help people do their best work, without making the process the point.

The next environment

I'm looking for good people, interesting problems and a role where I can take real ownership.

Product, client delivery and founder-side operations are the areas I know best, but I'm more interested in the work and the people than chasing a particular title.

I'd happily take a well-scoped role in a good team over a bigger title with the wrong fit.

Outside work

✈️ Aviation & flying lessons🚴 Cycling🗺️ Travel🛠 Startups & product building🏀 Basketball🏉 AFL🚗 Cars

Get in touch

Let's talk.

Based in Melbourne. Always happy to meet people working on interesting products, teams or businesses.

If you think there's something worth having a conversation about, get in touch.

Emailjay.stevens.digital@gmail.com

LinkedInlinkedin.com/in/jaystevens2

Email me →

Work with me

Selected project and fractional work.

I work with founders and product teams on short-term and fractional engagements, from shaping a new product idea to getting an existing product or delivery process back on track.

I bring eight years of hands-on product, delivery and customer experience from Launchpad, combined with AI-assisted workflows that let me take more of the work from idea to prototype myself. The value isn't the tools. It's knowing what to build, what to leave out and how to test the assumptions that matter.

Discuss a project →

Engagements

Three ways I can help

Zero-to-One Product Sprint

From idea to something you can see, test and confidently build.

Typically 6 to 10 weeks, shaped to the project

For

Founders and established businesses with a new product idea, opportunity or emerging customer problem.

Outcome

A clearer product direction, an interactive prototype and an evidence-backed plan for what to build next.

Can include

Problem and opportunity definition
Customer discovery and testing key assumptions
Product strategy and MVP priorities
User journeys and functional requirements
UX concepts and interactive prototypes
Initial technical scope, roadmap and commercial assumptions

Product & Delivery Reset

Find what's not working. Fix what matters. Get moving again.

Typically 2 to 4 weeks

For

Existing products or teams dealing with delivery bottlenecks, unclear priorities, customer friction or clunky operations.

Outcome

Practical improvements, fixed where the scope allows, and a clear plan for the rest.

Can include

A review of the product, delivery process or customer journey
Stakeholder interviews and workflow mapping
Prioritising the bottlenecks that matter
Hands-on fixes to processes, handoffs, requirements or customer experience
An action plan and next-quarter priorities

Fractional Product & Delivery Partner

Senior product and delivery capability, without another permanent hire.

Ongoing, part-time, with agreed scope and cadence

For

Founders and growing businesses that need experienced, hands-on support across product, customers and delivery.

Outcome

Real progress against agreed monthly priorities, with hands-on accountability rather than occasional strategy calls.

Can include

Leading a defined product or delivery initiative
Turning customer insight into product priorities
Coordinating engineers, designers and stakeholders
Building practical delivery systems and workflows
Decision-ready recommendations and ownership of agreed outcomes

I work independently and can bring in trusted design or engineering specialists where a project needs additional expertise. Every engagement is scoped around the work required.

Experience behind the offering

I spent eight years at Launchpad helping founders take digital products from early ideas through to launch, while also helping run and grow the business itself.

Through Startmate First Believers, I've added an early-stage investment lens to that experience, reviewing founder applications, pitch decks and business models. It's about looking beyond whether something can be built to whether the opportunity is worth pursuing, the evidence stacks up and the founding team is worth backing.

More recently, I've been putting that experience into practice with Velo Passport, building my own product using AI-assisted workflows to take more of the process from idea to prototype myself.

Start a conversation

Got something in mind?

Send me a short note about what you're working on and where you're stuck. I'll come back to you on whether I can help and what that could look like.

Email me about a project →

Active · pre-launch

Velo Passport

A digital passport for the places cycling takes you. Log a ride and it becomes stamps across the countries, regions, climbs and long-distance routes of Europe.

Role

Founder: product, design and AI-assisted build

Status

Pre-launch · design prototype

Research

Three rider conversations, June–July 2026

Started

During a year in Paris, 2025–2026

The idea came from a year riding around Europe. I'm building it end to end: discovery, product decisions, UX, visual design and AI-assisted development. This page covers what other riders told me, how the product changed as a result, and what is still unproven.

Evidence key Hypothesismy reasoning, not yet tested Rider feedbackheard from intended users Decisionadopted in the current spec

01The opportunityHypothesis

Cycling apps record rides. I wanted a record of where I'd been.

The apps I used captured the activity: GPS tracks, distance, training data. None gave me a lasting, visual record of the countries, regions, climbs and routes I'd ridden. My hypothesis is that other touring riders want that too, as a layer on top of the tools they already use rather than a replacement for them.

How I explain the positioning. This is a framing, not the result of a full competitive review.

Strava

Records rides, performance and social activity.

Komoot

Helps riders plan where to go next.

Garmin

Captures device, health and training data.

Wandrer, StatsHunters

Suggest some riders enjoy map-based completion.

Velo Passport

What I'm testing: remembering where you've been, and making it collectible.

Not a daily habit app

I expect riders to open it around trips, not every morning. So the measures I'd use are rides logged per rider, riders who come back with a second trip, passports shared, and purchases after a rider has seen their own map fill in. Session length doesn't matter.

Before a trip

Look at what's left to collect
After a ride

Add it and confirm the stamps
After a trip

Revisit the map and share it
Months later

Come back with the next trip

Expected rhythm, not yet observed.

02What riders told meRider feedback

Three riders, and not much agreement with my first idea

In June and July 2026 I showed the early EuroVelo Passport concept and landing page to three cyclists who ride in Europe. What they said is paraphrased from my notes and, for John, from an email he sent afterwards. It's a small qualitative sample, not validation.

Max

Has ridden Bordeaux to Biarritz

What came up

Hadn't heard of EuroVelo, even though he'd ridden part of a EuroVelo route.
More interested in practical trip help: GPX files, route information, places to stay, cafés and bike shops.
Stamps on their own weren't a reason for him to pay.

What it meant

I couldn't assume riders know the EuroVelo name. His planning needs were real, but they belong to a different job from keeping a record of rides.

Christian

Cyclist in Europe

What came up

Also didn't know EuroVelo as a route network.
Related more to places, regions and journeys than to an official route name.

What it meant

Reinforced a broader vocabulary: countries, regions, climbs and journeys, with EuroVelo as one collection among several.

John

Tours four to eight weeks a year in the Alps and Pyrenees

What came up

Liked how the landing page looked, but said his GPS history and photos already cover his memories.
Saw climb stamps shared back to Strava as a nice extra, not something he'd pay more for.
Spent about 50 hours planning a Pyrenees trip across guides, Strava, Garmin and his own research, and would value saving that time.
Questioned whether his age group of touring cyclists would pay for an app.

What it meant

Liking the presentation isn't the same as wanting the product. He also pointed to a second, separate opportunity in trip planning, and to price sensitivity in his segment.

The research surfaced two different jobs. I chose one.

Decision

In the first release

Remember where I've ridden

A map and stamps built from rides you've already done. Small enough to build and test on my own, and it's the problem I set out to solve.

Deferred on purpose

Plan a good trip, faster

Max and John described real planning pain. John spent around 50 hours planning one Pyrenees trip. A planner would need routes, stays, cafés and bike shops, which is a different product with far more data, content and support.

I didn't dismiss planning. I kept it out of the first release so I could test the collection idea properly, and I've noted it as a possible separate product to explore later.

03How the product changed

Four decisions that reshaped it

The spec is on its fourth version. It started as a EuroVelo route tracker with one promise: connect Strava and see what percentage of a route you've completed. These are the changes that moved it from there to the current spec.

From a EuroVelo tracker to a cycling passport for Europe

Rider feedbackDecision

Before

EuroVelo Passport: track progress along Europe's long-distance route network, starting with one route and one completion percentage.

Current spec

Velo Passport: five collections (countries, regions, climbs, EuroVelo routes and classic itineraries). EuroVelo is one of them.

Why

Two of the three riders I spoke to didn't know the EuroVelo name. Riders also cross countries, regions and climbs that sit outside the network. A passport built around places rather than one route brand fits how they described their riding. This broadens the product; it doesn't prove the broader version has a market.

In the design prototype

The Passport tab shows every collection side by side. EuroVelo sits alongside the others, with no special status.

Country

Region

Climb

EuroVelo

Classic

Adding a ride comes first, and nothing is matched silently

Decision

Before

Connect Strava once and the app quietly turns every ride into route progress. GPX upload as a fallback.

Current spec

Add a ride by typing a start and finish or drawing it on a map. GPX upload and a single-ride Strava import sit alongside. Every match is shown to the rider to confirm before a stamp is added. There is no automatic sync.

Why

Logging a ride is how a rider gets to the reward, so it should feel deliberate rather than happen in the background. Silent matches would make the passport harder to trust, and depending on one data source would leave out riders who don't use it.

Confirmation screen listing matched stamps with checkboxes — 1 · Review matches

→

Stamped screen after claiming — 2 · Stamped

In the design prototype

After a ride is added, the prototype lists what it found: here a country, a climb and a EuroVelo route. Each match has its own checkbox. Nothing reaches the passport until the rider taps Claim, and they can edit the selection first.

The EuroVelo row also shows decision 03: the route stamp is owned at Bronze after one ride.

Stamps are owned first and upgraded later

Decision

Before

Route progress shown as a percentage, so a partly ridden route read as unfinished.

Current spec

Riding 25 km of a route earns its stamp outright. Bronze, Silver, Gold and Complete are upgrades layered on top.

Why

A rider who has done a weekend on a route has earned something. Showing it as a gap would turn a memory into a to-do list. Tiers give a reason to come back without taking anything away.

25 km

First

→

10%

Bronze

→

25%

Silver

→

50%

Gold

→

95%

Complete

First stamp at 25 km ridden on the route; tiers by share of the route. Line collections only. Country, region and climb stamps are simply collected or not.

Rethinking the subscription

Rider feedbackHypothesis

Before

The written spec (July 2026): a free Explorer tier, an annual subscription and a limited founding lifetime pass. No monthly option.

Current direction

Let riders build their map and see it fill in first, then offer a one-off purchase. The price and exact rules aren't set.

Why

Older European touring cyclists pushed back on recurring payments, and John questioned whether his segment would pay for an app at all. Several described the product more like a travel diary or a scratch map: something you buy once and keep filling in. A product people open around trips may not suit a subscription.

This came from conversations, not a pricing test, and the written spec still shows the annual model. Whether riders will pay, and how much, is unproven.

04The model

One ride, tested against every collection

In the current spec, every ride becomes the same simple record: a start, a finish and the line between them. It doesn't matter whether it was typed in, drawn on a map or imported. That record is checked against each collection using the geometry that suits it, and the rider confirms the matches before any stamp is added.

One ride

A → B

Typed start → finishDrawn on the mapGPX fileSingle-ride Strava importNo automatic Strava sync. Bulk history import sits in Profile.

Classify → confirm → stamp

Countries Area Ride enters it Collected

Regions Area Ride enters it Collected

Climbs Point Ride passes within range Collected

EuroVelo Line Share of route ridden Owned, then tiered

Classics Line Share of route ridden Owned, then tiered

About 105 collectibles in the current spec. A single manifest defines them, so the total grows as collections are added and is never hardcoded.

05Design

Maps first. Passport second. Stats third.

The reward for logging a ride is seeing your Europe fill in, so the map leads. The app has three tabs: Home, Add and Passport. Profile and settings sit behind the header. The brand borrows from travel journals and old passport stamps rather than performance sport.

Home tab with map — Home
The map leads. Totals sit on top of it; the latest stamp and recent rides come after.

Stamps earned from a ride — After adding a ride
The stamps a ride earned, once the rider has confirmed them.

Passport tab with collections — Passport
Each collection and how much of it you hold.

Design prototype screens with sample data.

Palette

Forest#214E42

Cream#F6F2E8

Euro Gold#D4A64A

Slate#4A5568

Mark and stamp are kept apart

The shield is the brand. Circular seals are reserved for stamps the rider collects. An earlier circular logo was dropped because it blurred the two.

Velo Passport is an independent project. It is not affiliated with or endorsed by the European Cyclists' Federation or EuroVelo; the name is used only to identify the routes riders collect.

06Scope

What it deliberately isn't

Each of these was considered and left out to protect the core loop. Trip planning came up as a real problem in the research; it's out of the first release on purpose, not because it doesn't matter.

Trip and route planningTurn-by-turn navigationCafé and stay listingsSocial feedTraining analyticsAutomatic Strava sync

07The build

Built with AI, decided by me

Notion holds the product thinking, specs and decisions. GitHub holds the code. Claude Code does most of the building, inside DevOS, the workflow I put together that moves each feature through specification, design, build and testing, with review points where I make the call.

Claude CodeDevOSNotionGitHub

08Where it stands

What I've learned, and what I still need to prove

So far I have early qualitative feedback and developed product thinking. I haven't demonstrated sustained demand or willingness to pay yet.

Heard from riders

Two of three riders didn't know the EuroVelo name, so the product now leads with places rather than one route brand.
Planning a trip is a real and time-consuming problem. It's a different job from remembering one, and I've left it out of the first release.
A rider can like the presentation and still say the product isn't for him.
Touring cyclists pushed back on a subscription, which led me to explore a one-off purchase.

Not yet proven

Whether the broader Velo framing works better than EuroVelo alone, and for whom.
Whether riders keep logging rides and come back after a later trip.
Whether riders will pay, and whether a one-off purchase suits them better than an annual one.
How accurately rides can be matched across all five collections.
Whether riders can be reached without paid advertising.

Next milestone

Put it in front of pilot riders and see whether they log their rides, share their passport and pay for it, before building more.

I'd keep going if

riders share their passport without being asked, come back after a later trip, ask for more collections, or pay once they've seen their own map fill in.

I'd pause if

riders like the idea but don't log rides, or finding them depends on paid advertising.

Visit velopassport.co ↗

Concept · interactive prototype Nervous System Companion

Moss

A small companion for burnt-out nervous systems. Say how you're doing, get one thing to try that fits what you have left, then go and live your life.

Nothing to track. Nothing to keep up with.

Role

Product concept, UX, visual design and prototype

Status

Interactive prototype v3 · beta site drafted

User research

Not yet. Testing is the next step

Name

Moss is provisional

Moss is a product concept I took from a written brief to a working, clickable prototype and a beta landing page. It hasn't been tested with users yet, so this page covers the thinking, the decisions and the design, and what testing needs to answer.

Evidence key Hypothesisreasoning, not yet tested Decisionbuilt into the prototype No user feedback yet, so nothing is tagged as feedback.

01The problemHypothesis

People often know what helps. They can't reach it when they need it.

Under prolonged stress, a tired nervous system keeps running after the day is over. Small things cost more than they should, and most wellbeing advice assumes more capacity than the person has. The gap Moss is designed around is retrieval: in the middle of overwhelm, it's hard to remember what worked last time.

For

Adults under prolonged stress, persistent overwhelm or burnout who want everyday help coming down a gear, without needing a diagnosis. Also people in demanding jobs, carers and anyone going through a big change.

Not for

Diagnosing or treating any condition. Moss offers everyday wellbeing support. It isn't therapy, medical treatment or a crisis service, and it points people to real support when they need it.

North star

Within about a minute of opening the app while overwhelmed, the person feels understood and knows one manageable next step, or feels allowed to do nothing. Success often means closing the app.

02Product decisions

Less effort exactly when capacity is lowest

Screens from the interactive prototype, showing four important product decisions.

One suggestion, sized to what's left

Decision

Check-in asks how you are and how much you have in you: 20, 5 or 1 minute. Moss offers one thing to try. "That's too much right now" makes it smaller, more than once, until doing nothing is the suggestion.

Why

Menus of content ask for decisions when someone has the least capacity to make them. A single, smaller-on-request suggestion keeps the effort below what the person has.

Memory is only what you save, in your words

Decision

When something helps, the person can save it under the moment it's for: wired, overwhelmed, drained, afraid of going backwards. Next time they're there, it waits on the home screen as "Past you left you something".

Why

The core need is "I've felt like this before. What helped last time?" Their own words carry more weight than generic advice, and nothing is inferred or profiled.

Take something off the plate

Decision

Lighten the load starts from "What could we stop spending energy on?" Pick one area, like social plans, and get a kind way to let it go, including a short message to copy. Moss never sends anything.

Why

When capacity is low, adding another healthy habit can make things worse. Removing one demand is often the more useful step.

Nothing to keep up with

Decision

No streaks, scores, points or missed-day messages. Moss never needs looking after. Opening it twice a day or once a month are both fine, and the home screen always offers "I'm doing okay".

Why

A wellbeing app that creates obligations becomes one more thing to manage. The app should invite people to leave and get on with their lives.

03A core journey

Wired and tired, made smaller until doing nothing is the answer

One of seven journeys the prototype covers end to end. The person picks how they feel and how much they have in them. Each time they say "That's too much right now", the suggestion gets smaller, until rest is the suggestion.

→

→

→

→

→

04Seasons, not levels

Recovery isn't a ladder

My Journey offers four seasons. People pick the one that sounds like now, or more than one, and Moss changes what it offers to match. There's no quiz, percentage or reset: a hard weekend in Expand doesn't send anyone back to Shelter.

Shelter

Things are too much.

Small things to try

Cancel one thing
Eat something warm
Lie down without your phone

Reboot

You're remembering that you can influence your own life.

Small things to try

Step outside for five minutes
Choose what’s for dinner
Message someone you like

Rebuild

Capacity is returning.

Small things to try

Plan one thing, not five
Stop before you’re tired
Add back one thing you used to like

Expand

Life becomes bigger than recovery again.

Small things to try

Say yes to something new
Make plans with people
Notice what you want next

Underneath the seasons

Working lenses, not a clinical model

Resolution

Helping your system recognise that something is over and doesn’t need constant processing.

Resilience

Slowly widening how much stress you can meet, and how well you come back from it.

Recalibration

Changing the boundaries, priorities and old patterns that led to overload.

Reconnection

Returning to people, pleasure, identity, curiosity, work and life.

05Visual identity

The weather changes. The lantern stays lit.

The garden follows the weather in someone's life, not their worth. Storms never remove growth, and Moss never looks sick, sad or neglected. The prototype also compares a second visual direction, an indigo lantern refuge, for beta testers to help choose between.

Palette

Linen#F5EFE4

Lantern#F2C26B

Bark#8A5A2B

Dusk#1C1D33

Ink#2A2620

Type

"Yeah. Today was a lot."

That's too much right now

Newsreader for Moss's voice. Atkinson Hyperlegible, designed for low-vision readers, for every control. Quiet mode stops all motion with one tap.

06Where it stands

A complete prototype. No users yet.

Built

A stateful, clickable iPhone prototype covering onboarding, check-in, three quick tools, Field Guide and My Journey.
Seven end-to-end journeys, including "too much" stepping down to rest.
Six garden conditions and a second visual direction to compare.
A beta landing page with a sign-up that stores nothing on the site.

Testing needs to answer

Does someone understand what it's for without being told?
Does the companion feel reassuring or gimmicky?
Is the garden comfort, or a hidden score?
Does a note from past you help more than generic advice?
Do people come back when they need it, without being prompted?

The name, the companion and the artwork are provisional. Suggestions would be reviewed by clinicians before any launch, and success would be measured by useful exits and repeat use at moments of need, not session length.

Moss offers everyday wellbeing support. It isn't therapy, medical treatment or a crisis service, and it doesn't diagnose or give medical advice.

Concept · interactive prototype Energy Companion

Keeplight

A companion for the days your body has other plans. For life with energy-limiting illness, it sees the effort others don't, remembers who you are beyond illness, and expects nothing in return.

Nothing to track. Nothing to prove.

Role

Product concept, UX, visual design and prototype

Status

Interactive prototype · concept website

User research

Not yet. Needs testing with people who live it

Sibling of

Moss (Nervous System Companion)

Keeplight is a sibling concept to Moss, built on the same idea of a quiet companion but for a different situation. It has a working prototype and a concept website, and hasn't been tested with the people it's for. This page covers what makes it distinct and what testing needs to show.

Evidence key Hypothesisreasoning, not yet tested Decisionbuilt into the prototype No user feedback yet.

01The problemHypothesis

The part nobody sees

For people living with energy-limiting and fluctuating illness, including ME/CFS and Long COVID, much of the effort is invisible. Most health apps answer with tracking, scores and goals. Keeplight's premise is that being seen, and staying connected to who you are, matters more than managing a number.

What others seeWhat it took

A cancelled plan. Hours of hoping you'd manage it, the quiet maths of what it would cost, and a message rewritten four times.

A quiet day at home. Being bored out of your mind and unable to do a single thing about it.

Someone who's gone a bit quiet. Someone who still cares about all the same people and all the same things.

A good day. A good day you're trying very hard not to spend all at once.

02Same family, different job

Why Keeplight isn't Moss with a new skin

Moss helps someone come down a gear and choose a next step. For many people with energy-limiting illness, a next step isn't the point, and suggestions can read as pressure. So Keeplight keeps the companion and drops the coaching.

Moss

Respond to overwhelm, conserve capacity, choose one manageable step.

Keeplight

Witness and support, without measuring or coaching anyone's energy.

Shared

A companion that never needs looking after, no streaks, quiet mode, and message drafts the person sends themselves, or not at all.

Different

No check-in form and no suggestions sized to capacity. In their place: recognition, company, and a record of who the person is beyond illness.

03Product decisions

Being remembered, not reassured

Recognition, from your own words

Decision

On a clearer day, the person can tell Wick a little about themselves, one optional question at a time. On a harder day, Wick gives it back: it shows what they said, where it came from, and lets them forget it.

Why

Generic reassurance could be said to anyone. Being remembered is what makes support feel personal, and only repeating what the person chose to share keeps it honest.

Your lights stay on

Decision

My World holds lights for the people, loves, values and somedays that make someone who they are. They never fade, need tending or go out because the person has been unwell. Only they can take one down.

Why

Long illness can shrink a life to symptoms. The lights keep identity visible on days when doing any of it isn't possible.

Company without obligation

Decision

Wick can simply sit with you. Light the lamp, dim everything, or leave. Nothing to answer and nothing to keep up with. Coming back after days away gets a neutral hello, never a guilt message.

Why

For someone with very little energy, any request costs something. The companion has to be worth opening even when the person can do nothing at all.

Deliberately left out

Symptom logsEnergy scoresWearablesStep goalsStreaks and badgesExercise promptsDaily check-insA pet that misses you

04A core journey

Just get through today

A handful of taps, no questions about symptoms. Stay with Wick, have something familiar offered back, make one thing easier, or none of those. Every path can end with nothing done.

→

→

→

→

→

05Visual identity

Dusk, lamplight and very little movement

Wick is a slow tortoise who carries its shelter on its back and a lamp on top. It's fictional and honest about it. Two places, the Hollow and the Shore, and two shells. Quiet mode stills the scene, dims the colour and trims the words while keeping text easy to read.

Palette

Night#1C1A24

Lamp#F0C27A

Stone#C7BEB1

Paper#F1EADF

Ember#A2463A

Type

"You're still you. Even on the hard days."

Not today

Literata for Wick's voice, Atkinson Hyperlegible for controls. A different serif from Moss so the two never read as one product.

06Where it stands

To be shaped with the people who live it

Built

A clickable iPhone prototype: home, My World and its lights, things Wick remembers, Just get through today, message helper, comfort settings and a clear route to real help.
Two places, two shells and quiet mode.
A concept website that invites people to register interest, storing nothing on the site.

Testing needs to answer

Do people living with energy-limiting illness feel seen by it, or talked down to?
Is a companion welcome, or one more thing on a hard day?
Does having their own words given back help on foggy days?
Is it clear, without explanation, that nothing is being tracked?

Before it becomes anything more, it needs testing with people living with ME/CFS, Long COVID and other energy-limiting illness, and review by clinicians wherever it touches health.

Keeplight is not a medical device, a treatment or a crisis service. It doesn't diagnose, monitor or give medical advice.