ExigoTech-Beno/mikeoss
Azure AI Foundry-targeted fork with a Python FastAPI backend and AustLII / Australian legal plugins.
ExigoTech-Beno/mikeoss is a hard Azure pivot. The primary change is a new backend-agent/ Python FastAPI service intended to replace the TypeScript backend, built around Azure AI Foundry, Semantic Kernel, Azure PostgreSQL Flexible Server, Azure Blob Storage, and Entra External ID for authentication. The upstream TypeScript backend is left in place as a parallel track.
The standout contribution for evaluators not on Azure is the Australian legal plugin layer: austlii.py ships a Semantic Kernel wrapper over the AustLII SINO engine with a hand-built registry of roughly 90 database paths covering Commonwealth and state/territory courts and tribunals, and australian_legislation.py adds the 13 Australian Privacy Principles with a compliance checker. Both files are separable from the Azure plumbing.
What's in it
- Azure AI Foundry backend A parallel Python service designed to replace the upstream backend with a hosted Azure agent, built on Semantic Kernel.
- Australian legal grounding Plugins that pull from AustLII and other Australian legal sources, pointing at use inside Australian practice.
- Microsoft-stack opinionation The architecture commits to the Azure ecosystem end-to-end rather than staying cloud-neutral.
Direction
infrastructureintegration
Activity
Threads of work (detailed view)
ExigoTech-Beno wires Mike into Australia's case law
A fork built to answer Australian legal questions from primary sources - and to run inside a Microsoft enterprise tenant.
ExigoTech-Beno rewrites the backend in Python and wires in AustLII for Australian legal search
A single 1,800-line commit replaces the TypeScript backend with a Python FastAPI service targeting Azure AI Foundry, and ships two Semantic Kernel plugins covering roughly 90 AustLII database paths across Commonwealth and state/territory courts. The Azure plumbing is all-in; the AU legal data is portable.