I have a few questions for people who use OpenAI Assistants to communicate with website visitors or customers.
- Did you create the UI yourself or did you use ready-made solutions?
- Is communication with the end-user only text-based, or can the user, for example, send a file?
- Do you use one assistant or many and transfer the end-user between them?
- What languages do you support?
We have just released a solution that allows you to embed assistants on the website in the form of a widget: Free Chat Interface for OpenAI Assistants | OpenWidget and we are planning the next steps in development. Information from you will be very helpful.
There are open-source solutions for converting an Assistant into an embeddable widget such as this
Despite the lack of answers directly here in the thread, I received messages from several website owners, and a few also decided to test our solution and share their feedback - thank you very much for them!
I got answers the second question - I know that handling files from end-users is not perceived as a priority. This may only be useful in the case of a possible transfer of the conversation to a real human agent.
Two people that contacted me assumed that adding an assistant to a website would be more difficult than it actually was. For this purpose, we have created a short tutorial that shows what this process looks like: https://www.youtube.com/watch?v=VAm1RrDLpSg
thanks for all the messages, please keep it coming - directly here in the thread, to my dm or email.