A simple food tracker for Telegram that combines manual entry and AI-powered meal recognition.
You can keep a food diary directly in Telegram without installing the app. AI recognizes food from photos, saving time on entering and calculating calories; the paid version offers more recognition options. Manual entry is free. Reminders and a simple interface help you quickly add meals and see your daily balance.
Works directly in the messenger and allows you to track your food in a familiar environment, without installing separate apps or unnecessary steps.
A photo is converted into a dish card with nutritional information: the service determines the composition, volume, and key indicators faster than manual entry.
You can quickly add a dish using text or manually enter ingredients and portions if AI is unavailable or you need a more precise composition.
The service promptly reminds you to enter your meal plan, helps you maintain a routine, and avoid skipping meals to keep your food diary complete.
The user receives more recognitions and unlimited access to all features: history, statistics, and quick meal entry.
The team tested the trackers themselves, compared their strengths and weaknesses, and interviewed users with different lifestyles.
We defined the logic for AI recognition, manual input, tokens, and a premium mode to define the functional boundaries and structure of the future service.
We described scenarios: adding a dish, editing, statistics, reminders, and history, to ensure interaction remained simple and consistent.
We created wireframes, simplified the main actions, and designed screens in Figma to help users add meals more quickly.
We built the frontend and backend, integrated the AI model and token system, ensuring stable operation of the service in the Telegram environment.
We verified key scenarios, adjusted details based on testing results, and prepared the project for launch in approximately two months.
