You talk.
The record takes shape.

1. Context

NUTRI IA began with the observation that tracking food and macronutrients still takes considerable effort. In traditional apps, logging usually involves opening the app, searching for foods, selecting portions and adding each item.

The proposal simplifies this interaction: people describe what they ate, and the system organizes the record.

WhatsApp serves as the main input interface. A web/mobile application brings together history, goals and progress, connecting the convenience of messaging with a long-term view.

How can food logging become as simple as sending a message?

2. Problem

Daily food logging creates friction. Even someone motivated to track calories and macronutrients has to repeat steps after each meal. This effort is the barrier the project aims to investigate.

There is also an interpretation challenge: turning an informal message into structured information without treating an unspecified quantity as precise data.

“I had 150g of rice, chicken and some beans.”

The message includes foods, but not every quantity. The product needs to address these gaps so the ease of conversation does not compromise the clarity of records.

The clarification prompt shown on the phone is an interface exploration for this portfolio, not an already validated feature.

3. Role

I am the Product Designer and creator of NUTRI IA, leading the project from start to finish.

My role includes defining the problem and value proposition, designing the conversational experience, information architecture and the dashboard’s UX/UI.

I also explore the AI’s operating rules and the technical architecture needed to turn the concept into a functional product.

Building and deciding

As an independent project, I need to balance experience, technical feasibility and development speed. Decisions about capture, interpretation and storage are part of the experience even when they do not appear directly in the interface.

4. Evidence

The project is based on a behavioral hypothesis: instant messaging is already part of everyday life, while nutrition tracking tools require structured input.

Reducing the steps between “I ate” and “it was logged” could increase the frequency and consistency of tracking. This hypothesis still needs testing with real users.

What the MVP should measure

Meals logged per user per week; retention over time; time to log a meal; percentage of corrected entries; and frequency of dashboard visits.

These indicators should help assess whether the experience is simpler, the records are reliable and the history is useful to people.

This case study does not yet present validation metrics, interviews or demonstrated quantitative gains.

5. Decisions

01 / Capture where conversation already happens

I chose WhatsApp as the primary logging interface. People describe meals in a familiar channel, reducing the need to learn a new sequence of forms.

02 / Separate interpretation from calculation

AI interprets natural language and turns messages into structured data. Nutritional databases and business rules remain the system’s responsibility. This separation makes it possible to handle interpretation, calculation and review independently.

03 / Give each interface a clear role

WhatsApp for speed and logging. The dashboard for analysis and tracking. Charts, goals and visualizations do not need to compete with conversation for space.

04 / Preserve data beyond the conversation

Meals, calories, weight and goals need to exist as persistent, queryable data. The product’s memory cannot depend solely on an AI model’s conversation history.

6. Solution

NUTRI IA is a conversational assistant connected to a tracking platform. People can write naturally:

“For lunch I had 150g of rice, 100g of beans and 180g of chicken.”

In the proposed experience, AI identifies foods, quantities and context. The system structures this information, calculates calories and macronutrients, stores the data in the profile and responds with a summary.

Throughout the day, entries form a nutritional history. The dashboard would let people review daily intake, protein, carbohydrates, fat, previous meals, goals, weight and progress.

The product flow

Conversation → interpretation → data → history → visualization → learning.

These portfolio screens are a conceptual demo with illustrative responses and numbers. They are not connected to WhatsApp or an AI model and do not represent real nutritional calculations.

7. Results

The product is in development. The current outcome is a defined experience with two complementary interfaces: conversation for logging and a dashboard for understanding.

The project evolved from a chatbot idea into a product architecture with an independent layer of data, intelligence and history. WhatsApp becomes an input interface within that structure.

Next milestone

Validate the MVP with real users, observing logging effort, correction needs and consistency of use. Only then can the proposal’s impact be assessed.

The numbers shown in the prototype are interface data, not results for usage, retention or product effectiveness.

8. Lessons learned

The main lesson was realizing that AI can remove an interaction step: people do not have to organize their message like a form for the system to interpret it.

Another lesson was separating conversational memory from structured data. Meals and goals need to remain retrievable, editable and queryable over time.

This separation led to a broader understanding of the product and how its interfaces should complement each other.

A nutrition tracking platform whose primary input interface is a conversation.

The next challenge is to test whether the concept holds up in daily life, including ambiguous messages, changing quantities and corrections to a record.