Enhance Software Architecture Draft - Detail Machine Learning Pipeline

[dominic0df]

User story

1. As development team
2. I need to decide on a software architecture for the Machine Learning Pipeline with which our LLM will be trained
3. So that we achieve a common usage of technologies and our software is more flexible and easier to work on in pieces.

Acceptance criteria

• Research is done on how to use Named Entity Recognition (NER) to find key entities with which the data can be classifed into predefined categories
• The already existing architecture is more elaborate about how the Machine Learning Pipeline will work
• Rework on architecture draft
  • One software developer (preferably the Software Architect that created the Architecture Draft) presents the architecture draft with the POs to our industry partner (in a meeting with the POs and the industry partner)
  • Possible suggestions of improvement will be discussed with the team and then possibly the architecture documentation will be updated

Definition of done (DoD)

• Added only after week 5
• The same for all features
• Here goes the project specific part

DoD general criteria

• Feature has been fully implemented
• Feature has been merged into the mainline
• All acceptance criteria were met
• Product owner approved features
• All tests are passing
• Developers agreed to release