Watson

Team Members

Michael Watson Allan Yang Faisal Kherwish

Advanced course track

MLH prizes desires

Eleven lab headphones The raspberry pies

Requirements

The user has to have ollama and gemma3:4b downloaded currently for everything to work so if that's a hassle there is a video attached below

Inspiration

We wanted to make a better Siri/Alexa

What it does

Currently it acts as an AI question answering machine that is hosted locally on the user's machine.

How we built it

We used C# for the front end UI and Python for the backend API calls.

Challenges we ran into

Learning how to make an overlay using XAML, API connectivity issues

Accomplishments that we're proud of

We made a locally ran AI chatbot that will be used to become an AI assistant in the future.

What we learned

We learned how being full stack devs requires a lot of knowledge and practice to be good at

What's next for Watson

Taking our AI chatbot and using it to manipulate tasks that the user asks it. i.e if the user asks "I have a test on the 30th of this month" the AI will go into your google calendar and mark the 30th of whatever month thee user is in and state that there is a test and put the event on the calendar. Currently, the chatbot supports TTS when it reads the message back to the user, but we also want it to work passively in the background and turn on when the user says thee keyword "Watson" and then the AI will start listening. Hopefully we want to store context; users conversations when "Watson is not said" that way it can go back and look at what the user is asking in what context.

Built With

• c#
• python
• xaml