Context (high level)
-
Solo builder, early in the build.
-
Completed Speedruns; limited hands-on beyond that.
-
Goal: set a clean architecture now to avoid rework later.
What I’m aiming to do (kept intentionally generic)
-
Multi-tenant SaaS on Foundry with strict tenant isolation.
-
Ontology-centric reads; Actions for all writes.
-
Orchestrated background workflows (no customer data included here).
-
Analytics layer + optional AI-assisted recommendations that always require an admin approval step (no auto-execution).
-
External app ↔ Foundry via OAuth2 + REST.
Questions:
-
Dev Tier feasibility: In practice, can Dev Tier support strong multi-tenant isolation (e.g., per-tenant namespaces/workspaces, RLS patterns, separated service principals)? Or are there common blockers that typically require Enterprise?
-
Limits to plan around: Any Dev Tier caps that could constrain this approach?
-
Ontology scale (objects, relationships, versioning)
-
Actions throughput, idempotency/audit patterns
-
Automate / scheduled jobs / event triggers
-
AIP Logic/Agents availability or constraints
-
API/webhook rate limits or external integration boundaries
-
-
Design for minimal rework: If starting on Dev Tier, what early design choices help a future migration (naming/namespace strategy, per-tenant isolation approach, repo/port-adapter boundaries, approval workflows, logging/audit strategy)?
-
Learning path: Recommended training/certs for a solo builder beyond Speedruns?
-
Early-stage pathways: Any programs or suggested paths for projects that may need select Enterprise-only capabilities before fully committing?