No. Your data is only used to train your Cleva.Bot. Cleva.Bot does not add your data to ChatGPT. Rather, it adds the power of ChatGPT to your data. Cleva connects to OpenAI's API service to leverage ChatGPT's Large Language Model and AI capabilities. We do not share your data with OpenAI for LLM training. Your data is strictly your intellectual property and remains private and secure. Learn more about this at: https://openai.com/enterprise-privacy
A custom ChatBot is trained on your data, including your website contents and uploaded files, to gain a deep knowledge of your organisation, products and services. Combining this knowledge with the capability of a Large Language Model provides your business with a virtual assistant that can engage with customers in human typical conversation to accurately answers queries and make recommendations.
Each bot response is counted as a message. You have a monthly allocation of message credits based on your plan. ChatGPT 3.5 consumes 1 credit per message. ChatGPT 4 Turbo consumes 10 credits per message. You can add more message credits here at any time.
LINKS: URL's (all webpages); XML Sitemap; YouTube Video URLs; Google Sheets URLs;
UPLOADS: PDF; DOC; DOCX; TXT; CSV; XLS;
ChatGPT3.5 is better suited for the majority of chatbot purposes. ChatGPT 4 Turbo is superior for highly complex tasks. We'll help guide you to choose the best model for your needs.
No problem. You can add more credits here. Otherwise Cleva.Bot will collect users details to share with you so you don't miss any potential leads. Your credits will automatically refresh each month.
Yes! When your Cleva.Bot is ready you will receive a link to view and test your bot. The test page will also show the embed code for you to copy/paste to add/activate the bot on your website.
Yes! Your Cleva.Bot can be trained with any of 95 supported languages and your users can interact with your bot in any language.
Deployment turnaround is generally 2 to 5 days once your data has been received. This is variable depending on the volume and type of training data to be indexed; the extent of custom manipulation needed; and overall development/deployment demand.
Yes. This is encouraged and recommended. Each bot should be dedicated to a specific role. e.g. A Customer-Service bot and a Team-Training bot should be separate and independently trained for their intended purpose.
No and Yes. The monthly subscrption cost is the same, irrespective of the GPT language model selected. However, message credits are consumed differently. ChatGPT3.5 consumes 1 credit per message. ChatGPT4-turbo consumes 10 credits per message.