← All projectsAdvanced · About 12–16 hours
Azure OpenAI assistant over company data
The enterprise version of RAG: Azure OpenAI, Azure AI Search and a Data Factory or Fabric pipeline that keeps data fresh.
Good for: AI engineer, Cloud architect, Data engineer
What you will have at the endA chat assistant on Azure that answers from a refreshed data source, with secrets out of code and spend you can see.
- Azure OpenAI deployments
- Azure AI Search
- Data Factory or Fabric pipelines
- Managed identity
- Cost control
New to this? Do the toolkit setup first, then add these one at a time.
- 1
Create and inspect resources from the terminal.
$ brew install azure-cli
$ az login
- 2
- 3
- 4
How to show it to an employer
1Architecture diagram with every Azure service
2README with cost per 1,000 questions
3Short demo video
Stretch goals
Content filters and a blocked-request test
Bicep or Terraform for the whole setup
Compare with AWS Bedrock Knowledge Bases
Questions this prepares you for
Why Azure OpenAI over the public API for an enterprise?
How do you keep the index fresh?
How do you control spend?
Practise them in a mock interview Next projectNatural-language to SQL assistant→