Azure AI Development
Budget: ₹600 – ₹1,500 INR
We are looking for an expert Azure AI Developer to build a functional Proof of Concept (POC) / Demo for an internal AI Assistant within a HIPAA-regulated healthcare environment. The goal is to demonstrate a "Source of Truth" engine where natural language queries about business KPIs return consistent, governed data from a Databricks Lakehouse.
Key Technical Stack:
Platform: Azure AI Foundry (formerly AI Studio) & Microsoft Fabric.
Data: Databricks (Unity Catalog, Gold Layer tables), Azure Data Lake Storage (ADLS).
Governance: Microsoft Purview (for PHI tagging and RBAC).
Orchestration: Semantic Kernel or AutoGen for Multi-Agent workflows.
Integration: Deployment to Microsoft Teams or a custom Web App.
Scope of Work (The Demo)
Multi-Agent Architecture: Build an "Orchestrator" agent that routes requests to specialized sub-agents (e.g., Financial, Clinical, Marketing).
Data Consistency Logic: Implement a solution that ensures different phrasings (e.g., "How many people started treatment?" vs. "How many people received [Billing Code X]?") return identical results by mapping to the same verified Databricks "Gold" tables.
Governance & Security: Demonstrate PHI masking and role-based access using Microsoft Purview and Azure RBAC, treating the agent as a standard user identity.
Citations & Transparency: The AI must provide citations linking back to the specific Databricks table/source for every answer provided.
Deliverables
Functional Demo: A live environment (or clear deployment script) showing the agent in action.
Technical Architecture Diagram: High-level visual of the data flow and agent orchestration.
Required Experience
Proven experience building Agentic AI on Azure.
Deep understanding of Databricks Unity Catalog and SQL-based RAG.
Experience in HIPAA-regulated environments (understanding PHI security).
Ability to explain complex architectural decisions to technical leads.
Key Technical Stack:
Platform: Azure AI Foundry (formerly AI Studio) & Microsoft Fabric.
Data: Databricks (Unity Catalog, Gold Layer tables), Azure Data Lake Storage (ADLS).
Governance: Microsoft Purview (for PHI tagging and RBAC).
Orchestration: Semantic Kernel or AutoGen for Multi-Agent workflows.
Integration: Deployment to Microsoft Teams or a custom Web App.
Scope of Work (The Demo)
Multi-Agent Architecture: Build an "Orchestrator" agent that routes requests to specialized sub-agents (e.g., Financial, Clinical, Marketing).
Data Consistency Logic: Implement a solution that ensures different phrasings (e.g., "How many people started treatment?" vs. "How many people received [Billing Code X]?") return identical results by mapping to the same verified Databricks "Gold" tables.
Governance & Security: Demonstrate PHI masking and role-based access using Microsoft Purview and Azure RBAC, treating the agent as a standard user identity.
Citations & Transparency: The AI must provide citations linking back to the specific Databricks table/source for every answer provided.
Deliverables
Functional Demo: A live environment (or clear deployment script) showing the agent in action.
Technical Architecture Diagram: High-level visual of the data flow and agent orchestration.
Required Experience
Proven experience building Agentic AI on Azure.
Deep understanding of Databricks Unity Catalog and SQL-based RAG.
Experience in HIPAA-regulated environments (understanding PHI security).
Ability to explain complex architectural decisions to technical leads.