Lifestyle & NutritionSuper App.
70+ personal variables, wearables and evolutionary algorithms. Meal plans, recipes and shopping adapt to habits and feedback.
Profile variables
Condition–ingredient relationships
Ingredients
Recipes
People change.The plan follows.
A personalised nutrition platform needed to bring together biological profiles, scientific evidence, tastes and everyday life. The project addressed the entire journey: data collection, plan construction, recipes, shopping and adaptation through new data and feedback.
- Evolutionary algorithms
- Optimisation
- Personalisation
Inside the app.
The person, at the centre.
The home screen brings together goals and the daily plan. It lets people follow planned activities and share how their journey is going.
Every piece of information matters.
Short questionnaires gather habits, preferences and personal information. The profile grows over time and informs how the plan is personalised.
A plan that fits your life.
Days, meals, recipes and quantities become a plan available in the app. Preferences and feedback help revise the proposal and adapt it to everyday life.
From plan to shopping.
Recipe ingredients feed into the shopping list. This view connects meal planning with the purchasing and delivery options designed for the project.
A plan that evolveswith the person.
The project combines a multidimensional user profile, two knowledge maps and evolutionary algorithms. The result takes shape in an application connecting plans, recipes, shopping and feedback.
The foundations of personalisation
Personal profile
Over 70 variables collected through mini-quizzes, user-provided information and wearables describe goals, habits and relevant conditions.
- Biology and lifestyle
- Preferences
- Goals and routines
Evidence map
Nutritionists and biotechnologists organised over 5,000 relationships between conditions and individual ingredients, drawn from evidence and guidelines.
- Increase
- Avoid
- Use caution
Food and recipe map
Over 6,000 ingredients and 2,000 recipes are linked to energy, carbohydrates, protein, fat and more than 20 other nutrients.
- Foods
- Recipes
- Nutritional composition
What we designed.
Scientific knowledge provides context, optimisation explores combinations and software manages the experience. Integration matters because a profile change must consistently affect the plan, recipes and ingredients.
Organise profiles and knowledge
Mini-quizzes and wearables contribute to profiling. Two knowledge maps connect scientific findings with the nutritional characteristics of foods and recipes.
Optimise the plan
Evolutionary and genetic algorithms explore combinations of recipes and quantities, considering profile constraints, food pairing, preferences and everyday practicality.
Update the experience
The application connects plans, recipes and shopping lists. Progress, new information and feedback feed the recalculation of proposals and quantities.
Profile + two knowledge maps
Evolutionary optimisation + feedback
Plan, recipes and shopping
How the plan is built and updated
Personalisation brings together nutritional constraints, preferences and what is practical in everyday life.
Define the person’s constraints
Profile information becomes criteria for constructing proposals.
- Meal structure
- Breakfast, lunch and dinner can be organised across the week according to the person’s habits.
- Preferences and diet type
- Tastes, food choices and goals guide the selection of combinations.
- Practical recipes
- Recipe frequency, number of preparations, ingredients, preparation time and available kitchen tools become explicit constraints.
Explore combinations
Evolutionary and genetic algorithms search the space of possible combinations.
- Evolutionary scenarios
- Evolutionary iteration explores a space of millions of combinations, favouring those that better fit the profile and plan constraints.
- Food pairing
- Pairing logic helps construct combinations that also consider taste and practicality.
- Plan and recipes
- Ingredients and recipes are associated with quantities, with energy and nutrient calculations and dose adjustments at gram and milligram level.
Close the loop through use
The plan can change when data, habits or feedback change.
- Monitoring and feedback
- Weight, physical activity, symptoms, mood, new laboratory values and adherence to the plan can contribute to the updated profile state.
- Plan adaptation
- The engine recalculates recipes and quantities when changes or difficulties emerge, keeping proposals linked to objectives and constraints.
- Shopping list and purchase
- Recipe ingredients generate a shopping list connected to the in-app purchase and home delivery journey.
Personalisationis an evolving system.
Explore how the platform connects a person to scientific knowledge and updates the plan through new data.
Understand the person beyond the goal.
Over 70 variables describe characteristics, habits, preferences and relevant conditions. Mini-quizzes, entered data and wearables feed the profile.
- Goals and lifestyle
- Tastes and routines
- Information and feedback
A multidimensional profile that can be updated.
Two maps, one scientific foundation.
Evidence is linked to conditions in the profile; foods and recipes are organised by nutritional characteristics.
- 5,000+ condition–ingredient relationships
- 6,000+ ingredients
- 2,000+ recipes
Structured knowledge to guide optimisation.
Explore millions of combinations.
Evolutionary algorithms search for combinations of recipes and quantities that fit the profile, nutritional constraints and everyday life.
- Energy, macro- and micronutrients
- Meal structure and variety
- Cooking time and equipment
A personalised plan with recipes and a shopping list.
New data restarts the cycle.
Progress, adherence and new feedback update the context. The system recalculates proposals while keeping the plan, quantities and ingredients consistent.
- New profile data
- Feedback on the plan
- Updated recipes and shopping
A proposal that evolves with the person.
The engineering challenge is making the whole cycle work
Profiling, scientific knowledge, combination search and the application work on the same plan. When new data arrives, the system updates its proposal and keeps recipes, quantities and ingredient lists consistent.
Personalisation is an optimisation cycle: new data is reflected in recipes, quantities and shopping.
The choices the system needs to coordinate
- How can nutritional goals, tastes and recipe variety be balanced?
- How should the plan be recalculated when routines and feedback change?
- How can the plan, ingredients and purchasing journey stay aligned?
What it makespossible.
- Plans and recipes built around the person’s multidimensional profile.
- Proposals updated from new data and feedback.
- An integrated journey from personalisation to the shopping list.
Scale what makesyour service unique.
Want to scale a highly personalised service?
Tell us about your project
Precision Genomics for rare diseases.
From a VCF file to drug-related evidence. A platform connecting genetic variants, conditions and scientific sources.