Confluence Trigger to Automate Vectorizing

• Confluence will trigger when a page has been edited, a HTTP request will be send to API Gateway/OCI Functions
• OCI Functions will either write to OCI Streaming/Data Prepper/OCI OCI OpenSearch or directly apply logic

Steps

1. Create an API Gateway and deployment
2. Create Automation rule in Confluence using the API Gateway endpoint
3. Create OCI OpenSearch Clusters. Use: https://github.com/bobpeulen/oci_opensearch/blob/main/oci_opensearch_rag_auto.ipynb 4.Create OCI Data Science ntebook and model deployment
4. Create OCI Functions deployments

1. Create an API Gateway and deployment

• Create an API Gateway and deployment with PUT/POST. In this example, we use "confluence" and "updatepage" as paths.
• This example used no authentication, but you might add.

2. Create Automation rule in Confluence using the API Gateway endpoint

• Add the API Gateway deployment full path in rule

  {
  "page_id":"{{pageidsmart}}",
  "page_url":"{{pageurlsmart}}"
  }
  

3. Create OCI OpenSearch Cluster

• Create an OCI OpenSearch cluster, version 2.15, in public subnet, in same VCN as API Gateway.
• Create a OCI Data Science notebook in same VCN, private subnet + NAT gateway
• Follow these steps: https://github.com/bobpeulen/oci_opensearch/blob/main/oci_opensearch_rag_auto.ipynb. This notebook will help in deploying in RAG

4. OCI Data Science

5. OCI Functions

• Create repo/project and log in Docker

  docker login -u '[namespace]/OracleIdentityCloudService/bob.peulen@oracle.com' iad.ocir.io
  password = Auth token
  

• Create function

  fn init --runtime python confluencex
  cd confluencex
  

• Change the func.py and requirements.txt

• Test function. Deploy and invoke function. Default values should be printed (in logs) and returned.

  fn -v deploy --app app_bp
  fn invoke app_bp confluencex