Inspiration

The best personal assistants are proactive. They don't wait for you to ask them for help, they take the initiative and can tell you the hard truths. On the other hand, assistants like Siri or Alexa only do things for you when you ask them.

What it does

Waylon observes your behaviour and proactively gives you helpful suggestions. It is connected to your calendar, wearables, etc, and can take actions on your behalf.

How we built it

We built an event-driven system. The assistant is triggered by events like alarms, calendar events, predicates on wearable data, etc. When triggered, it gives the user a call and has a conversation. The user gets the proactive suggestion, and can then follow up with any requests to the assistant.

Challenges we ran into

We wanted to dynamically set the first assistant message and the system prompt, but the 11labs apis don't support that yet. Also the assistant often got interrupted by background noise and didn't finish saying the helpful suggestion. It would be nice to be able to disable interruptions for the initial message. We worked around the hard-coded first message limitation by creating many assistants.

Accomplishments that we're proud of

We're proud of having made progress on giving the leverage that people with personal assistants have to everyone. Currently personal assistants are something that only afluent people can afford.

What we learned

It was great to see how easy it is to build a conversational agent.

What's next for WAYLON.AI

We'll add integrations with more apis so that the assistant can provide more helpful suggestions.

Built With

Share this project:

Updates