WFT Chatbot (NEST and NEXT)

This is Langchain implementation using NestJS for backend and NextJS for frontend.

Installation Notes

Frontend

First, run the following command to go to the client directory:

cd ./client

Now run the following command to install the dependencies:

npm i

Now set the environment variable in .env file.

NEXT_PUBLIC_SERVER_URL=(server url)
# 5001 is the socket server port
NEXT_PUBLIC_SOCKET_SERVER_URL=<server_url>:5001

To run the development server:

npm run dev

Open http://localhost:3000 with your browser to see the result.

Backend

First, run the following command to go to the server directory:

cd ./server

Now run the following command to install the dependencies:

npm i

Now set the environment variables in .env file. Some of the environment variables are left as default:

PORT=(port to run the server on)
SOCKET_PORT=(socket port to run socket on)
MONGODB_URL=(mongodb connection url)
JWT_SECRET_KEY=(jwt secret keys)
DEVELOPER_EMAIL=(developers email separated by comma)
OPENAI_API_KEY=(open ai api key)
EMBEDDINGS_MODEL_NAME=text-embedding-ada-002
PERSIST_DIRECTORY=./lancedb
CHAT_MODEL_NAME=gpt-4
CHATBOT_TO_CONNECT_INFINITY_SERVER_STRING=(base64 generated string)
VECTORSTORE_COLLECTION_NAME=(collection name to infer by the ai)
VECTORSTORE_URL=(url to the vectorstore)

To run the development server:

npm run start:dev

The development server will now run on the localhost with the port defined on .env file.