Our first hackathon: How did it go?

Head of Data Science @MewsSystems; Former PhD student in CS Machine Learning @Matfyz #dataengineering #learningfromdata #machinelearning #javascript #python

At Mews, we spend a lot of effort to make our APIs well documented and easily integratable with other solutions. On Friday, we wanted to test it against chatbot implementations, so we gathered a group of developers and technological enthusiasts and challenged them to build chatbots on top of our platform.

We have partnered with Microsoft and they have provided us with some nice spaces and catering. At 10am, we kicked off with an introduction talks about the technology that the teams could use: Microsoft’s Azure Service Bot, Dialog Flow Designer by BooAI and our Mews APIs.

We began with introduction talks about Azure Bot Service, Dialog Flow Designer (by BooAI) and MEWS APIs.
We gave a quick walkthrough talk about our products.

We did not need to worry about getting hungry. Throught out the day, there was a plenty of meals and snack available and, most imporantly, coffee.

We partnered with Microsoft who took care of the event space and catering.

Around noon, we gathered all the ideas and let everyone decided what they want to work on and formed the teams accordingly.

Prabash presenting his idea of a chatbot.

Every team had their own meeting room available where they could really focus on working on their ideas, building the prototype and finalising the presentation.

There was no developer in buddy team, yet they managed to build a working implementation using Dialog Flow Designer and Luis.ai.
The winning mluvii team is figuring out the optimal dialog flow for their chatbots.

During the afternoon, we have been checking on the teams progress and trying to help them with the technological and design aspects of their projects.

After hours of work, we have invited everyone back to the main room and gave them 10 minutes to present their ideas and answer the questions from the judges.

At the end of the day, we have gathered to see the final presentations. Matt, our CEO, joined us later in the evening to see the presenations and help us choose the winning teams.

When reviewing all the ideas and presentations, we had quite some discussion, but in the end, we agreed on the final ranks and decided to merge 2nd and 3rd place.

  • 1st – mluvii team
  • 2nd/3rd – buddy team
  • 2nd/3rd – four team

mluvii team has presented a guest end-to-end chatbot. Being placed on top of hotel’s website, it gives the guest the direct access to support, checkin/out, and ordering additional products. They have used their existing solutions deployed to Azure.

Buddy team came up with a room service bot. It’s main purpose would be help the guests with their requests and complains about the room they’re staying in. For the hotel staff, it’d generate the analytics and give them better overview of the state of their rooms and services.

Four team tried to give their chatbot even broader specification. A concierge service available to both guests and hotel staff allowing them to ask for restaurant or another place recommendation and assisting them in the hotel lobby in general.

Thank you, everyone, who helped us with this event and also the participants! It has been amazing experience. We have both enjoyed it a lot and also learnt a great deal of ideas and feedback what can be improved next time.

Head of Data Science @MewsSystems; Former PhD student in CS Machine Learning @Matfyz #dataengineering #learningfromdata #machinelearning #javascript #python
Share: