EnterpriseFeatured2026

CMIS AI Decision Support

AI system in CMIS that identifies interrupted-in-treatment patients, with a privacy-preserving AI assistant for Power BI dashboards.

My role · Software Engineer

About this project

AI decision support built for a client-management health information system (CMIS) at Palladium: Make It Possible.

Interrupted-in-treatment: an AI system that identifies patients who have interrupted treatment so follow-up can be prioritised
Dashboard assistant: an AI model accompanies the Power BI dashboards and answers users' questions
MCP + DAX: the assistant reads data through a Model Context Protocol tool that generates DAX queries from the question and returns matching results
No database access: the model never connects to the database, only to the query tool
Privacy by design: processed, aggregated data is returned so personally identifiable information stays out of responses

Keep exploring

School Management Template

Venda Khona

End of case study