AI Chatbot Starter

This AI Chatbot Starter is designed to help developers find the information they need to debug their issues.

It should answer customer questions about the products or services specified.

Getting Started

Now follow the steps below to get the chatbot up and running.

Documentation (provided as a list of web urls in the config.yml) can be ingested into your Astra DB Collection. Follow these steps:

Obtain your OpenAI API Key from the OpenAI Settings page.
Create a config.yml file with the values required. Here you specify both the list of pages to scrape, as well as the list of rules for your chatbot to observe. For an example of how this can look, take a look at either config.yml.example_datastax, or config.yml.example_pokemon.
Create a .env file & add the required information. Add the OpenAI Key from Step 1 as the value of OPENAI_API_KEY. The Astra and OpenAI env variables are required, while the others are only needed if the respective integrations are enabled. For an example of how this can look, take a look at .env_example.
From the root of the repository, run the following command. This will scrape the pages specified in the config.yml file into text files within the output folder of your ai-chatbot-starter directory.
```
PYTHONPATH=. python data/scrape_site.py
```
From the root of the repository, run the following command. This will store the embeddings for the scraped text in your AstraDB instance.
```
PYTHONPATH=. python data/compile_documents.py
```

If you have Docker installed, you can run the app using the following command:

Build the docker image using the following command:
```
docker build -t docker_aibot --no-cache .
```
Run the docker image using the following command:
```
docker run -p 5555:5555 docker_aibot
```

You can test an example query by running:

python scripts/call_assistant.py "<your_query_here>"

Alternatively, you can run the app normally using the following steps:

Install the requirements using the following command:
```
pip install -r requirements.txt
```

Run the app using the following command:

uvicorn app:app --host 0.0.0.0 --port 5555 --reload

You can test an example query by running:

python scripts/call_assistant.py "<your_query_here>"