Euphoni

Inspiration

Businesses face high customer support costs and inefficiencies due to manual handling of customer queries. This leads to longer response times, dissatisfied customers, and missed growth opportunities.

What it does

We present a Voice bot customer support system that leverages natural language processing and AI to offer real-time, automated, and highly scalable personalized assistance. The voice bot efficiently resolves customer queries, reduces response times, and optimizes support costs.

How we built it

Mixture of claude api, vocode, langchain

Challenges we ran into

-Issues with twilio account made real time call impossible -trouble with ui, lack of front end experience on team -not enough time to flesh out fully working prototype

Accomplishments that we're proud of

-UI almost complete -iterated out way through the structure of the prompt, used xml structure to find optimal prompt that works.

What we learned

-Speech API toolings can be made better: Even though LLMs are sufficiently good at holding dialogue, the voice interface is only evolving but the applications are wide and presents huge market opportunity for disruptive innovation from call centers to car voice assistants