L’Assiet
Culturally aware meal planning built around households, local food and shared routines.
Plan the week. Reduce the mental load. Keep food human.
Explore the system ↓Outside the job description
Some problems won’t leave me alone.
These projects begin with ordinary friction I can’t quite ignore.
They’re where I explore behaviour, culture, systems, AI and new business models—turning observations from everyday life into products and systems tangible enough to test.
Outside the job description
Some problems won’t leave me alone.
Problems I have noticed around me, explored deeply and translated into systems that can be tested.
Styling platform
Personal styling, capsule wardrobes, outfit visualisation, wardrobe organisation and culturally relevant recommendations.
Make more of what you already own.
Explore the system ↓
REVIV
Keep → upcycle → resell → donate → repurpose or recycle → discard responsibly.
What’s its best next life?
Explore the system ↓
Innowear
Moving production knowledge out of paper, WhatsApp and people’s heads—so information can travel with the garment.
Less searching. Better decisions.
Explore the system ↓
Product system 01
L’Assiet
A living meal-planning system designed around the person’s health goals, preferences, routines and changing reality—not around perfect compliance.
How might healthy eating fit around real life—instead of asking real life to fit around a meal plan?
People rarely need more recipes or nutrition information. The harder problem is turning intention into everyday decisions while accounting for health, culture, portions, preferences, available food and a changing schedule.
Feedback from step 05 returns to the profile and plan ↺
01 · Profile
Understand the individual
Build a living picture of goals, health context, preferences, lifestyle and constraints.
02 · Personalise
Translate context into a plan
Turn that information into meals, portions, nutrition targets and realistic weekly structure.
03 · Support
Make decisions easier
Recipes, substitutions, portions and practical guidance live in one system.
04 · Reflect
Capture real-life behaviour
Use lightweight check-ins and progress signals rather than demanding perfect tracking.
05 · Adapt
Feed learning back
Adjust what comes next based on what worked, what didn’t and what changed.
Product thesis & positioning
Personalisation framework
Nutrition logic
Meal-planning system
Recipe framework
Behavioural feedback loop
AI interaction model
Product architecture
Product system 02
Styling intelligence
I designed and tested the intelligence an eventual styling platform would need—before treating an interface as the product.
How might we teach AI to style a real person—not an idealised version of them—using the wardrobe they already own?
The wardrobe is the starting asset. Shopping comes later, and only when a specific addition meaningfully unlocks more of what the person already has.
Human expertise becomes testable system intelligence
01 · Frame
Define the real problem
Move beyond outfit generation toward wardrobe potential and style evolution.
02 · Research
Study people and wardrobes
Lifestyle, climate, proportions, occasions, comfort, taste and aspiration.
03 · Codify
Translate styling judgment
Turn expertise into principles, guardrails, methods and AI instructions.
04 · Prototype
Test on real clients
Style resets, capsules, colour direction, wardrobe edits and recommendations.
05 · Learn
Interrogate failures
Body integrity, garment fidelity, relevance, practicality, age and visual consistency.
06 · Systemise
Make quality repeatable
Translate learning into prompts, workflows, training and quality standards.
07 · Productise
Define the platform
Use validated styling intelligence to shape the experience and requirements.
Styling methodology
AI training framework
Client research system
Visual prototyping system
Wardrobe intelligence
Onboarding logic
Product requirements
MVP testing framework
Product system 03
REVIV
A Mauritius-based circular wardrobe experiment designed to help people declutter, keep, transform, resell, donate and circulate clothing more intentionally.
What is the best next life for the clothes already in your wardrobe?
Most solutions begin after someone has chosen to sell, donate or discard. REVIV begins one decision earlier and connects the entire choice around a garment’s next life.
Community becomes the research environment
01 · Activate
Community
Build a trusted environment around high-quality pre-loved clothing and better wardrobe decisions.
02 · Observe
Real behaviour
Watch how people declutter, price, list, buy, donate, ask questions and make decisions.
03 · Learn
Evidence
Capture friction, trust signals, language, repeated requests and successful or abandoned transactions.
04 · Interpret
Validated insight
Separate anecdotes from patterns before allowing findings to influence future product thinking.
Brand & positioning
Community model
Content & education
Listing standards
Donation ecosystem
Local resource hub
Research framework
Insight handoff system
Product system 04
Innowear Cashmere
A connected operating model for a traditional Mauritius-based knitwear company—preserving craftsmanship while connecting information from customer order to delivery.
How do you modernise a 40+ year-old knitwear business without losing the expertise that made it successful?
I mapped the business as it actually operates, extracted the specialist knowledge inside it and identified where software could remove friction without replacing judgment or craftsmanship.
Start with reality. Prove the architecture on one garment.
01 · Observe
Map the real business
Private sales, wholesale, sampling, production, inventory and administration.
02 · Follow
Trace real orders
Customer → design → plan → yarn → production → finishing → delivery.
03 · Extract
Capture expert rules
Measurements, grading, yarn consumption, production decisions and costing.
04 · Diagnose
Find the friction
Paper records, repeated calculations, handoffs, stock visibility, pricing and tracking.
05 · Bound
Define technology’s role
Digitise information and repetition; preserve judgment, relationships and quality decisions.
06 · Design
Connect the business
Orders, products, plans, yarn, inventory, production, costing and delivery.
07 · Prove
Start with one garment
A basic V-neck becomes the pilot from specification through actual margin.
Information · calculations · tracking · history
Craftsmanship · judgment · relationships · exceptions
Business process map
Digital operating model
Product architecture
Knowledge-capture framework
Inventory system
Costing & pricing engine
Production tracking
Phased MVP roadmap