SaaS AI Chatbot Platform
What Distinguishes It Our platform features some unique elements:
- Voice AI Calling: We partnered with Bland.ai to enable voice interactions, allowing users to communicate with their bots via voice calls, which adds a personal dimension to digital conversations.
- Custom Knowledge Base: Sales representatives can upload their own knowledge bases, allowing chatbots to access a specific and relevant source of information to accurately respond to inquiries.
- Personalization: Each sales rep has the option to customize their chatbot to reflect their individual personality and style, aiming to make digital interactions feel as natural as talking to a friend.
- Visitor Tracking: Our platform monitors website visitors and records their chat histories, effectively serving as a built-in CRM within the chatbot, which simplifies follow-ups and personalization efforts.
- Secure Access: We have established a strong authentication system to ensure that only authorized sales reps can manage and access their chatbot settings.
Exploring the Technology This project was not solely focused on the final product; it involved navigating through some advanced technology:
- Rapid Prototyping: Our initial success was the development of a prototype that could efficiently integrate a knowledge base with a Large Language Model (LLM), allowing the chatbot to retrieve and process information effectively.
- Front-End Development: Using Vercel's NextJS framework, we designed a user interface that is both intuitive and visually appealing, making interactions with the chatbot enjoyable for users.
- Tech Stack: We utilized Vercel for hosting, providing us with the necessary flexibility and scalability. For the AI capabilities, we depended on OpenAI's GPT-4 for its exceptional natural language understanding and generation. The backend was supported by Postgres, with Langchain JS assisting in our interactions with the vector store, enhancing chatbot responses with relevant contextual information.