Inspiration:
As a dietitian, after I discovered that If there is a strict diet, there is no sustainability, the mission of Simple Diet is formed. Many people are struggling to fit their life in a diet list. When they eat the foods that are not in the diet list, they are losing confidence and they feel unsuccessful and unhappy. Moreover, they are quitting to lose weight. When we make some researches, We've seen similar results in many studies. Only 3 of 20 people are continuing to diet at the end of one month. Furthermore, Most of people who lost weight by strict diets can not maintain kilo that they've lost. They are gaining back. We have to integrate high-calorie foods to provide a sustainable diet process, and we have to take into consideration of social activities, modern life style. We developed a Turkish chatbot that works on whatsapp. And the results of users inspired us for developing it in English too. We would like to reach people from all around the world.
What it does:
The users can start to conversation with Simple Diet by writing ‘’Hi’’ on Facebook Messenger. Simple Diet needs to calculate energy to be taken daily. For this reason, Simple Diet asks body weight, age, height, physical activity status at the first conversation. Simple Diet doesn't bother the user to count calories, but it calculates the calories in the background and gives them portions according to calorie that calculated at first conversation. After this, the user can ask its meal with the name of the foods that user want to eat. And in couple seconds, Simple Diet chatbot replies with the amount of consumption advice. The users can ask breakfast, lunch, dinner or snack. User has to write the name of meal.
How we built it:
We are using webhooks to get messages from Facebook messenger. We use our custon nodejs server for nlp functions. We send response messages via Facebook messenger send api
Challenges we ran into:
Recognition of food name was challenging. Because it’s not an optional chatbot. It’s 100% conversational. That’s why we have to match the foods that user write with the foods on our database correctly.
Accomplishments that we're proud of:
• In Turkish version of the chatbot on WhatsApp , we saw that 75% of the users are continuing to diet. • We increased the food name recognition rate from 67% to 78%. • The users lost 1,1-2,2 lbs per week. • English version of this chatbot is working!
What we learned:
We learned how we can use facebooks api to create a chat bot.
What's next for Simple Diet:
• To increase the food name recognition rate 95% • To have more than 1 million foods in our database. • Voice recording recognition.



Log in or sign up for Devpost to join the conversation.