Feature issue : #1706

Write a design proposals for Gen AI data ingestion workflow using :

• Gitlab pipeline as data ingestion scheduler
• OpenSearch as vector DB provider
• AWS lambda to run ingestion script with access to the database
• AWS for infrastructure (this design may include GCP GKE reflexion also)
• Langfuse as test dataset storage solution
• Reuse as much as possible existing python tooling : tock-llm-indexing-tools
• Optional Ragas for evaluators

Design should be reviewed and approved before starting any development to be sure that we are developing in the right direction.