Have a different question and can’t find the answer you’re looking for? Reach out to our support team by emailing us, and we’ll get back to you as soon as possible.
- Do you offer a free plan?
- Yes! You can use Chatterdocs free with one chatbot, three sources, and up to 50 requests per month.
- What is a source?
- A source is used to train your chatbot. For example, webpage or file (PDF, Word, RichText, Markdown, MP3, etc.).
- How is it different from ChatGPT?
- ChatGPT is trained on a wide amount of data, while Chatterdocs is trained on your data. We use the same large language model technology behind the top AI tools (e.g., OpenAI, Claude, etc.).
- What is a request?
- A request is a question to your chatbot. Questions are logged to track usage.
- How is my data stored? Is my data safe?
- Your data is vectorized and stored in a Postgres database hosted with Supabase. We use row-level security for all data and do not store copies of your files. The vectorization process allows AI to search through your content for relevant answers based on the questions asked.
- Can you prevent prompt hacking?
- Our app only uses your content to provide answers, and will not pull from Chat GPT's general knowledge.
- What data do you store?
- We currently store your content, custom prompts, and visitor questions.
- Do you integrate directly with apps like Notion?
- Yes, we currently integrate with Notion and plan on integrating with other apps soon.
- What do you use for your natural language processor?
- We are using OpenAI's Chat-GPT API, but have built the app to use other vendors in the future. Our hope is to eventually support multiple AI models to prevent vendor lock-in.
- Do you or OpenAI retain and train future versions of GPT with your data?
- OpenAI only retains context for 30 days and does not train their models with the context you provide.