Implement a Natural Language to Requirements translator

[mlhaufe]

Goal

We need cathedral to integrate with common business chat and collaboration platforms like Slack, Teams, and Zoom Chat

There are existing and potential customers who write arbitrary requirements in this chat channels as a form of TODO list.

• Implement LLM service (Create LLM Service #350)
• Translate natural language into the requirements grammar
• Store the requirements in the DB
• Define a chatbot in one or more of the mentioned platforms
  • Slack (Git repo)
  • https://api.slack.com/docs/apps/ai
  • https://slack.dev/node-slack-sdk/tutorials/local-development/

[mlhaufe]

Some references and resources:

• https://learn.microsoft.com/en-us/microsoft-copilot-studio/fundamentals-get-started?tabs=web
• https://jd-bots.com/2022/02/12/connect-a-microsoft-azure-bot-to-slack-channel/
• https://dev.botframework.com/
• https://www.microsoft.com/en-us/microsoft-copilot/microsoft-copilot-studio
  • Introduction Video
  • The pricing is prohibitive...
  • $200/month for 25,000 messages. Request and response counts as two messages. as does talking to any backend api. So a Slack connector would be at least 4 "messages" for every call. I guess they have to pay for those new Nuclear Power Plants somehow