Table of Contents / Содержание
Main purposes of the service
- Reduce the cost of maintaining a contact center
- Provide round-the-clock customer service
- Automate typical tasks
How it works
-
There is a telegram chatbot for customers (Customer Bat) through which they communicate with LLM, which uses a knowledge base about a particular business to provide accurate information. If desired, customers can sign up in a queue where they will wait for an agent (business representative).
-
There is a telegram chatbot for agents, through which they can get information about a waiting client with a summary of his communication with the bot in order to quickly find out about the client's problem. There are also special functions for the superagent, which is set via a script in the Generative Backend.
-
The link between the two bots is the Generative Backend. It is necessary for generating content and working with the database.
The service is designed for small businesses, since it is expected that the entire service, including LLM, will be deployed on a local machine owned by the business. Without the need to use third-party services and computing power. This is primarily due to the high prices of third-party products.
The project is based on a micro-service architecture
There are 4 main services -
Main functions of each of the services
customer_bot_speed.mp4
agent_bot.mp4
- Run all services according to the README for each project.
Roman Neronov:
- email: roman.nieronov@gmail.com / roman.nieronov@mail.ru
- telegram: @Romiconchik