We’d like the agent to figure out how relevant a workload is—whether it’s critical, niche, or part of a planned deployment—based on tenant telemetry. This means understanding M365 dependencies for each post. For example, if Purview is in use, posts related to Purview should get higher priority than if it’s not being used. Categories can also be established based on a tenants license assignments and M365 services usage.
Acceptance Criteria
The agent can access tenant data via MCP or Graph API:
/subscribedSkus
ServicePlan mappings
User counts per SKU
Optional usage reports (if enabled)
The agent computes relevance using rules such as:
License penetration (% of users enabled)
Usage levels (if available)
Workload criticality (Teams, Exchange, etc.)
Deployment state (not used, partially deployed, fully deployed)
Relevance classifications are included in enriched output.
The agent explains the reasoning behind relevance scoring.
We’d like the agent to figure out how relevant a workload is—whether it’s critical, niche, or part of a planned deployment—based on tenant telemetry. This means understanding M365 dependencies for each post. For example, if Purview is in use, posts related to Purview should get higher priority than if it’s not being used. Categories can also be established based on a tenants license assignments and M365 services usage.
Acceptance Criteria
The agent can access tenant data via MCP or Graph API:
/subscribedSkus
ServicePlan mappings
User counts per SKU
Optional usage reports (if enabled)
The agent computes relevance using rules such as:
License penetration (% of users enabled)
Usage levels (if available)
Workload criticality (Teams, Exchange, etc.)
Deployment state (not used, partially deployed, fully deployed)
Relevance classifications are included in enriched output.
The agent explains the reasoning behind relevance scoring.