📝 geoffdavies-dt frees Mike from a single AI provider 3 commits 2mo ago infrastructureintegration draft
Mike can now point at Azure's hosted OpenAI service - or other compatible endpoints - instead of only OpenAI's public API.
geoffdavies-dt/mike
Single targeted PR from Dovetail: makes the OpenAI provider configurable so self-hosted deployments can point at Azure OpenAI or any compatible Responses endpoint.
geoffdavies-dt (Dovetail) has one recorded change: adding OPENAI_BASE_URL and OPENAI_AUTH_HEADER env vars to backend/src/lib/llm/openai.ts. The default behavior is unchanged; setting both vars routes requests to Azure AI Foundry or a compatible proxy. The PR was merged back to main as PR #1.
That's the fork's entire contribution. Narrow scope, clean execution, no open issues.
What's in it
- Azure OpenAI support Point Mike at your own Azure-hosted GPT deployment instead of the default OpenAI service - useful for teams whose AI stack already lives in Azure.
Direction
integrationinfrastructure
Activity
Threads of work (detailed view)
geoffdavies-dt frees Mike from a single AI provider
Mike can now point at Azure's hosted OpenAI service - or other compatible endpoints - instead of only OpenAI's public API.
Two env vars make the OpenAI provider point at any compatible Responses API endpoint
geoffdavies-dt (Dovetail) added `OPENAI_BASE_URL` and `OPENAI_AUTH_HEADER` to the OpenAI provider so mike can talk to Azure OpenAI or any other Responses-compatible endpoint. About 20 lines of code, fully backwards-compatible.