Messenger Bot. Event posters. Personal Assistant.
What it does
Messenger Bot capable of understanding photos input and save it accordingly to its type: either an event or a contact.
How we built it
It uses the Mix.NLU API from Nuance to interact by text with the bot. We're also using Microsoft's cognitive API to translate photos to text. Events and contacts are actually saved in a database so the user can ask for it later. We also use the Messenger Platform, Microsoft Azure to host the bot and Node.js.
Challenges we ran into
It was hard to settle to one precise AI idea at first. We also had to filter text received from the Microsoft cognitive API to only keep english words and numbers/dates.
Accomplishments that we're proud of
We were able to complete the product we wanted to make in time.
What we learned
As we were all beginners of AI before coming to the hackaton, we learned a lot (thanks to the workshops) it the field and we were able to implement some AI API in our project to make it a smart bot. We also learned about Nuance's Mix.NLU API which is really powerful to understand variations in sentences that utlimately have the same meaning.
What's next for EventAI
Adding other features to the bot so it becomes a better personal assistant. We also want to train our model to understand a larger number of sentences.