About the Project
Inspiration
A few days ago, as I was walking to the university along Font Blvd, I saw an elderly lady waiting at the stop for the 57 Muni bus. She didn't know how to check the bus arrival times on her phone and asked me when the next bus was coming. Interestingly, I've encountered this same lady multiple times at the same stop around the same time. This incident made me realize that many people, like my own grandparents and even my parents, struggle with using smartphones effectively. My grandparents, for instance, have smartphones but primarily use them just for making phone calls. If they ever need to troubleshoot something or view pictures I send, I always need to guide them step-by-step over the phone. Not everyone has someone available to help them like that.
This inspired me to create an AI-based phone service that provides companionship, assistance, and information through simple phone calls, accessible even to those not comfortable with smartphone technology.
What I Learned
Through this project, I learned how critical user-centered design is, especially when developing solutions for seniors or individuals who aren't tech-savvy. I explored integrating conversational AI platforms, telephony APIs like Twilio, speech-to-text, and text-to-speech technologies. Additionally, I gained deeper insights into building accessible technology, ensuring ease of use through natural language interactions.
How I Built the Project
I used Twilio's telephony API to handle incoming phone calls and integrated it with a Google's Gemini model to provide a natural, conversational experience. Gemini 2.0 supports speech-to-speech conversations which allows the model to understand the user's emotions and other small but important voice features and respond accordingly. I also integrated APIs to fetch live data like Muni bus arrival times and general web search information, making the AI genuinely useful for daily tasks.
Log in or sign up for Devpost to join the conversation.