OneCenter

Struggling with bad customer service? OneCenter's your solution! Our AI call center agent tackles support issues, and learns & personalizes interactions, all while saving you time & money

Comment

Inspiration

Despite the emergence of new technologies, call center customer service is still poor, and getting a lot of complaints, this is a worldwide issue but the issue of having to contact a call enter is greatly frowned upon in Nigeria West Africa and if possible would rather be avoided. We decided to make use of these technologies to help better customer service for businesses and make customers happier

What it does

OneCenter is an innovative AI-powered call center solution that centralizes multiple companies' call centers. Using cutting-edge AI and machine learning, it provides smart and timely responses to customer queries over the phone, delivering personalized support and eliminating long wait times. OneCenter revolutionizes customer service, making it faster, more efficient, and cost-effective for businesses.

How we built it

OneCenter is built using large language models, combined with Convex's database vector search feature, to implement a RAG System, interfaced with a SIP protocol, enabling customers to make voice calls over IP (VoIP), on our web UI. The languages used for this solution are Python, JavaScript, and Typescript.

Challenges we ran into

One of the challenges we ran into while developing, was the fact that the langchain vector store convex database had not been implemented, so we had to implement a custom vector store for convex db from scratch. Another Challenge we had was setting up the SIP call feature. Getting a hang of the platform we used (Radisys) took a lot of effort, and was ultimately a stressful process

Accomplishments that we're proud of

We built a fully functional MVP, in the space of approximately 1.5 months. We heard about the Convex hackathon at the beginning of February.

What we learned

  • Effective time management
  • Gained more experience in the agile methodology
  • JavaScript ^_^

What's next for OneCenter

As a startup business

  • Beta Testing
  • Fund Raising
  • Product Development (re-iteration)
  • Product Launch
  • Market Expansion

For the dev Team

  • Exploring more advanced querying and data modeling capabilities of Convex; we noticed that Convex has yet to implement more sophisticated filters and filter syntax for the vector search feature. This is something we look forward to as developers. We really love the new feel that convex db brings regarding vector search and how databases are interacted with.

Built With

Share this project:

Updates