I have a question about some greenfield architecture and wondering how to best achieve this. Using .NET 7 and wanting to host on Azure with cloud-native services if possible. We (2 people) are creating a chatbot that summarizes conversations from WhatsApp using OpenAI. We have the following problem regarding the archicture design: Twilio sends a post request whenever someone sends a whatsapp message to a specific receiver to our API endpoint. The message gets put on a message queue (service bus) and this triggers an Azure function to process a response to the user or end the conversation We see that people send whatsapp messages right after each other, sometimes with delays of up to 2 minutes, these messages trigger functions and are being processed as a conversation on its own. We need to aggregate these messages into one single message to draft a response and if the conversation is ended a summarization.
I am trying to think of ways we can create this without the need of distributed locks or timers. Anyone have some ideas how to best design this?
Recent Announcements
Continue the conversation
Join the Discord to ask follow-up questions and connect with the community
C
C#
We are a programming server aimed at coders discussing everything related to C# (CSharp) and .NET.