OWNED PRODUCT · ACTIVE DEVELOPMENT
EcclesiaNexus
A greenfield church operations platform designed around explicit module boundaries, multi-tenant ownership, cloud/edge operation, and real on-site workflows.
- Role
- Architecture & product owner
- Status
- Active development
- Architecture
- Cloud core + edge agent
- Stack
- Go · React · PostgreSQL · SQLite
Implemented Today
The repository is well beyond an architecture sketch. It contains a Go cloud API, PostgreSQL persistence, a Go edge agent with SQLite caching, React 19 web applications, Docker-based deployment, CI workflows, and implemented domain modules including WAMS, roster/scheduling, People, Check-Ins, and Groups. Notification infrastructure and operational/admin capabilities are also present.
Cloud / Edge Boundary
The cloud server owns authoritative multi-tenant data. The edge agent provides a local execution and cache boundary for church-site workflows where local availability matters. Cloud-to-edge synchronization is deliberately modeled as a boundary rather than sharing a database between deployment contexts.
Domain Boundaries
The system separates global users, workspaces, workspace membership, and domain people identities instead of collapsing them into one table. Cross-module identity decisions are documented and implemented through explicit bridges, with workspace isolation enforced in data access and authentication paths.
Verification
The project includes automated Go tests, frontend type/build checks, CI, integration QA, and gated real-PostgreSQL end-to-end verification. The Groups flow, for example, verifies cross-module person linkage, capacity and leader rules, attendance idempotency, directory scope, and module gating against a real database.
Still In Progress
This remains an active product, not a finished commercial system. Additional church-management and commercial capabilities are tracked separately from implemented modules. I keep that distinction explicit so roadmap intent is not presented as shipped functionality.
AI-assisted Engineering
I own problem framing, architecture, boundaries, contracts, decomposition, review, and acceptance. AI agents accelerate detailed design, implementation, testing, and documentation within those constraints.
What this demonstrates
Greenfield backend architecture, multi-tenant domain modeling, cloud/edge integration, explicit module ownership, automated verification, and the ability to use AI as an engineering force multiplier without outsourcing architectural responsibility.