In contrast to conventional search methods, PandaChat leverages GPT's comprehension of various languages found on the Internet. Hence, even if your documentation is indexed in English, it can provide responses in the language of the inquiry. Presently, querying in a different language than its training data may yield less effective results, as it may struggle to locate the most relevant context from our embeddings index. However, we are actively working on enhancing this capability!

We employ Open AI's GPT to process your documents and address your inquiries. Our team has fine-tuned the models and prompts to extract pertinent information from your documents, striving for truthful and factually accurate responses. However, it's important to note that GPT is a relatively new technology and our service is still in its early stages. Consequently, it's crucial to understand that the answers provided may not be entirely reliable, although they are correct in the majority of cases (approximately 90% of the time). Additionally, we want to emphasize that in accordance with our Terms, we do not assume any responsibility for the factual accuracy or truthfulness of the answers generated by DR. Panda, PandaChat's chatbot. When using PandaChat chatbots, please proceed at your own risk.

No, there are no limitation regarding uploading files. With PandaChat you can engage with more files at the same time. PandaChat supports different files, some of them are: The document formats we support are PDF, DOC/DOCX, TXT, XLSX, PNG, JPG, MP4, and PPT. We do not have limitations when it comes to pages for a single chatbot. We do not have limitations when it comes to mbs in size for a single chatbot. All limitations are based on the pricing plan you choose. Check pricing here.

If you are dissatisfied with your paid subscription to PandaChat, kindly inform us within 30 days by sending an email to [email protected] We will gladly assist you in processing a refund.

It may sound complex, but here's a concise overview. Our system utilizes OpenAI's embedding and ChatGPT APIs, along with vector databases to store our index. The documentation we ingest undergoes a cleaning process and is divided into smaller sections, labeled by their source. Next, we employ the GPT embedding API to generate a vector representation for each section, which is then stored in our vector database index. When a user poses a question, we convert it into an embedding and conduct an advanced semantic search to find the closest matches to the query. We retrieve the most relevant sections and include them as context, along with the original question. Using the ChatGPT API, we generate a response in markdown format, which is subsequently converted to HTML and presented to the user.

Certainly, our AI has been explicitly designed to alert users when it is unable to find relevant information in the given document. As a result, all the AI's responses are solely based on the information extracted from the source documents. Hence, if you trust the document or article, you can have equal trust in the AI's response!

All data is securely stored in cloud storage, guaranteeing both the safety and privacy of the information.

We have Stripe as our payment processor, enabling us to accept payments made through credit and debit cards.