Inspiration

Covid-19 has changed the educational world. Online based learning is becoming one of best option. Zoom.us is the most popular tool for meeting agenda. Unfortunately, Indonesians face problems about online based learning, i.e. connection issue, expensive internet access, and low-spec mobile device . But the meeting agenda should be done using zoom. Not all people can join the meeting. We, as programmers, want to help their to solve the problem. We make a bot to do some great assistant's work.

What it does

Metabot is a zoom meeting assistant bot that can automatically summarize all of the meeting conversation. It gets the audio data, convert to text, then summarize the data to become a final note. This bot can lists all participants based on the name too. We built the front end web, so everyone can use this bot.

How we built it

We built metabot using 3 components. 1) the zoom-web sdk, this website is used to capture all of meeting activities after selenium login on this website. 2) Metabot, this is a backend that handle request and response from the frontend web and integrate it with selenium. 3) Metabot-fe, this website is intended to users if they want to call the bot to do automatic text summarization and list all participants. This website also show the analytic every meeting agenda.

Challenges we ran into

Do Automatic process using selenium on Zoom website was very hard, We got the problem when using the zoom API to create a meeting. It must past through so many ways. Integrating between frontend-backend-database-selenium was a long and tired process because sometimes, other members were not yet doing their task.

Accomplishments that I'm proud of

Finally, we did and solved all of that problems except using zoom API to create a meeting. We learned so much about selenium, nuxt.js framework, google speech api, flask, etc. We know that there are similar tool like we built but we proveide the front end web for users. So, this is user friendly and easy to use. But the main accomplishments that we're proud of is the impact of our bot. We hope that this tool can be used by Indonesians so they can take benefit from the bot.

What I learned

We were not only learning about hard skill, we learned about soft skill too. How we managed our ego and emotion during the hackathon. Blaming person is not a good solution. Never surrender until the end of time and work the best is our key.

What's next for Metabot

We want to fix bugs and and make it more proper for production. We want to add some features such as face emotion and recognition.

Share this project:

Updates