I am looking to try Momen to create a app that has API calls across lots of different lists for users. After getting the response from the API it appends that response to a value that can be grabbed from another DB on the platform.
I can’t find it in the documents, but is there a way to pass values from a DB through an API dynamic values similar to bubble? Each user will have unique values that need to be passed through the API for response to be worth anything.
Also, any idea when trying to test with “https://api.openai.com/v1/chat/completions” it just gives back “hostname can’t be null” but removing “api.” does give a response.
If I understood correctly, you are trying to send API out with dynamic values. But that’s almost the only way to send API calls.
If you want to do it on the backend, you would setup an actionflow to do that. I.e. fetch data from db, and then use that in an API call’s parameter.
In terms of api.openai.com, that’s a bug, we had a proxy that is no longer needed. Will get it fixed soon. Though I’d recommend you use the AI functionality directly.
Is there any documentation on passing variables from action flow into the API portion? I noticed that the AI is using Momen tokens and seems limited to being a chatbot rather than to gather insights and storing the data. The documentation doesn’t seem to have any reference on saving the response to a DB (other than the user messages DB)?
you can use AI inside actionflow. make sure it is async. The momen AI functionality is a lot more than a chatbot. It’s a full agent. You can checkout the youtube livestream https://www.youtube.com/watch?v=LMOPqANQmy0
you need to setup your API first, then everything should be natural. It doesn’t work like “gimme a url and do everything else adhoc”. you need to properly set up API first.
Each component is only responsible for its own thing, so in order to achieve what you need, you need a solid grasp of both actionflow and momen AI, as well how to deal with the momen database.
In Momen, Actions stored in the database can be configured for buttons on the frontend or added as nodes to insert data in the backend ActionFlow. Additionally, if the conclusions from AI analysis need to interact with external data, you can configure custom outputs, which can then be used in ActionFlow (ActionFlow needs to be set to asynchronous mode). Below are two relatively complete examples of using APIs and AI: