Description

Key responsibilities

• Ingestion pipelines: Build and maintain Python pipelines that ingest identity and metadata from

Entra ID (Microsoft Graph API) and other selected sources, including pagination, throttling, retries,

delta queries and idempotent loads.

• Graph modeling and queries: Implement graph schemas and write Cypher queries for users,

nested groups, roles and resources, including access-path and entitlement queries.

• DataHub integration: Run DataHub metadata ingestion through the GMS APIs and ingestion

recipes, handle schema changes and re-ingestion, and configure metadata synchronization

between components.

• Entra ID security mapping: Map Entra security groups and roles to access attributes used

downstream by the semantic layer.

• Pinecone Nexus context curation: Produce structured schema payloads and curated context, with

metadata and permission attributes needed for filtered retrieval.

• Custom connectors: Develop required connectors, including a Service Desk Knowledge Base

connector.

• Use cases: Build and deploy at least two simple initial use cases on the context layer.

• RBAC POC support: Build the data orchestration from Entra ID, GCP Cloud Asset Inventory API and

AI MCP access layers to identify over-provisioned, stale and overlapping roles.

• Quality and operations: Add logging, error handling, tests and monitoring, and write runbooks

and technical manuals for what is built.

• Collaboration: Pair with the Lead Architect and RadNet engineers, take part in reviews and sprint

ceremonies, and flag blockers early.

Must-have skills

• Python: Strong production-grade Python for data pipelines, APIs and automation (not just

scripting)

Key Skills

Education

Bachelor's degree