Yellfi

• 

  Creatiing a Yell

• 

  Pledging for a Yell

• 

  Reading a Yell's Details

Inspiration

When we travel to somewhere, we always check various sources to find help or to get a recommendation. We may not always have a connection who knows the area. So we wanted to build a solution to let people find the recommendations they need easily.

What it does

First of all, Yellfi asks the user for a location. This is the area that the recommendation will be based around. Then the user writes what they need. This can be a recommendation for a good restaurant or someone who has an extra voucher that will expire soon. If the others check the Yellfi around the location, they will see the post (Yell) and answer to the Yeller. The Yeller can also pledge for something in return.

How We built it

We used Microsoft Bot Framework. For Geosearching, We are using ArcGIS (address to GeoLocation) and using PostgreSQL for storing yells and make geospatial queries.

Challenges We ran into

There isn't many examples for Microsoft Bot Framework ( webpack - NodeJS ) so we struggled when we tried to deploy it to production. Getting user's locations is not accurate with adress to lat,long , and we are writing a webview screen for it. However, the most difficult part in the future will be sorting through spam posts. For now, a Yell (post) is active for just 12 hours and has 10 km radius.

Accomplishments that I'm proud of

Even if the chatbot environment isn't as big as other mobile apps, especially in the terms of development, we created a location based chatbot. This is a unique solution that we haven't seen a similar app before.

What We learned

We learned so much technical details about Chatbots, Messaging techs, and NodeJS but the important thing is building a chatbot has some challenges that not just in terms of development. Building Chatbot is creating a persona. So this persona has to be able to connect the users to each other using our solution. This was a new challenge that we've taken on.

What's next for Yellfi

While we will be developing the technical side, the main next step for us is on the commercial section. We will be looking for more use cases for Yellfi, especially in the business space where this app can be used in partnership with other companies that may not be able to reach to their client base. A recommendation service and sponsored recommendations can be a pathway we may want to explore.

Built With

• awslambda
• awsrdspostgresql
• microsoftbotframework
• node.js
• webpack