Inspiration

All of us are very active messengers' users. Many of our events are arranged in social networks. But sometimes we lose an invitation, forget about it or do not know anyone who would like to join us. The desire to solve this problem has brought us to WeeGo.

What it does

Imagine, you chat with your friends everyday, and often receive offers to get together, or you want to go somewhere but don't have a partner. In this situation WeeGo is a useful smart observer, who remembers your desires and looks for the best solutions to satisfy them. It detects actions, places and time, sends you invitations and receipts. You can also pay for the ticket with Telegram payment system. Finally, you may use aggregated by bot data in an android application where you will find beautiful interface to discover new people and new places. For long-distanced rides we provide users with integration with Skyscanner: discuss and track travels and flights easier than it ever was!

How we built it

We use NLP with requests to Google Places to fetch place, location and datetime of an action. Also, we generate previews with Bing Images. Our server has been deployed on AWS EC2. We have combined a lot of different APIs and SDKs, including Skyscanner, MyQR, Compreno, Google Places, Bing Search.

Challenges we ran into

First, we have attempted to parse Facebook data for another idea. But after some time we have faced privacy troubles and decided to switch to new topic. The general challenge to us was not only create an app, but create a useful app. If we talk about challenges of sponsors, we ran into ImaginBank chatbot challenge and Skyscanner challenge.

Accomplishments that we're proud of

We have made a very good and friendly team with a nice separation of tasks. We improved our skills and results since last Hackathon, we have tested a lot of API and successfully combined them. In our humble opinion, the biggest accomplishments of our work are usability and stability of the product.

What we learned

We have learned many new technologies. The funniest one is building QR-codes based on pictures (thx, BIENE). Also it was our first time to build application with monetisation system and payments. We have studied new APIs: Skyscanner, Bing, new aspects of REST API. It was also our first experience in developing a chat-bot and nlp.

What's next for WeeGo

Among other things WeeGo is a way of data collection. Our next goal is to expand the service, finish data acquisition and train generative adversarial networks to add more options in text recognition. For the marketing part, we would like to cooperate with commercial organisations, create discounts for our customers and prepare data for future smart engines.

Share this project:
×

Updates