The Build Team.
Private working room. Four people. Enter the build-team password to continue.
The Build Team.
This is not the intranet. It is the room where the four of us plan it. What we've read, where we are, which questions are open, and what we're doing next — kept in one place so we don't have to reconstruct it every meeting.
Four people. Four surfaces.
We each own a different surface of the same problem. When a question comes in, this is the first thing to check to know whose column it lives in.
Deon Graham
Identity, hosting, backend, and the Google Workspace surface.
Melissa Walker
The shell, the design, adoption, and the covenant docs.
Dr. Jacqueline Lawler
Records, information architecture, and the standards each space follows.
Valencia Johnson
Cadence, decisions, blockers, delivery.
Decisions phase — no code
The Build Spec, the Grow Guide, and the Design Review are all in hand. The Design Review pushed back hard on custom-building the board portal and the LMS. Before a line of application code is written, the chair needs to answer three questions (see the one-pager) so buy-vs-build is settled.
- No launch at or near the 26 September retreat. Show the map, not the door.
- Nothing on this room's page is real auth. Do not paste anything sensitive here.
- No church data goes into AI coding prompts. Synthetic data only.
The intranet in four documents.
Read in order: the one-pager for the chair, the Build Spec for what was proposed, the Grow Guide for the discipline that keeps it small, and the Design Review for the independent read that redirects the plan. Word format, printable.
Buy or Build?
One-page decision for the chair. Start here.
Build Spec and Claude Code Prompt
The proposed architecture, modules, and tiered build plan.
How To Grow It — a decision guide
The discipline for what to add, and what to say no to.
Full Design Review — six disciplines
The independent review that redirects the build. The one to argue with.
Six answers stand between us and code.
Each row has an owner in this room. Move to "In progress" when someone starts working the answer; move to "Answered" and add a note when it lands.
How many intended users lack an Embassy Google account, and are we bringing everyone into the tenant (Cloud Identity Free)?
PendingIndependent conflict route — a named person and an alternate, adopted by board resolution, who receives reports about the apostle, an Executive Pastor, or the CMCO.
PendingReal annual software budget for board portal, LMS, and the shell's dependencies.
PendingTree test before any code — thirty to forty volunteers, unmoderated.
PendingIs Planning Center in the building? If yes it is the source of truth for people and campuses, not a new database.
PendingCaptions and transcripts on the four credentialing videos, before anyone enrols.
PendingWhere this is trying to go.
The design review's revised sequence — deliberately slow, deliberately gated on decisions rather than dates. Nothing here is a promise; it is the shape we're aiming for.
-
Now → 25 September
Answer the six questions. Draft the Space Owner Covenant with Dr. Lawler. Board resolution naming records custodian, safeguarding lead, triage owner, and the independent conflict route. Build nothing.
-
26 September (retreat)
Show the map, not the door. One slide on how existing tools connect and where the index will live. No URLs, no logins.
-
Weeks 1–3 after
Run the tree test. Caption the credentialing videos. If a shell is going to be built, this is where the shell starts — with the PEDM room only.
-
Mid-October
Pilot with the PEDM board alone (roughly 12 people). One full monthly cycle: packet, acknowledgements, meeting, minutes, decisions log.
-
November
Prove cycle two. Fix what broke. Do not expand until two consecutive cycles run clean.
-
December–January
Open credentialing ECCI wide, anchored on an Alliance Sunday. Ship the small universal module: directory + "am I serving this Sunday."
-
February onward
EMD in full from a seven-field intake, then one campus at a time — each gated on a signed owner and a live Access Steward.
The intranet does not need to be good to survive. It needs to be visibly, boringly consistent for six months. What we decide in this room decides whether the organisation gets to trust the next new thing we ship.