Prepared for Joe Occhipinti, Jireh Billings, Charlie Dinh and Natalia Kovaleva · 29 October 2026

Finish the move. Retire the old tools. Close the list.

The platforms run the company's day now. This is the plan to put them on ANC's own ground by the end of November, turn off the tools they replaced, and finish the year with nothing still depending on me. Every number below was read off the live systems today.

1 Stand2 Move3 Retire4 Open5 Next6 Calendar7 Needs

01

Where things stand

The platforms are the system of record live

In the last eight weeks the service team raised 706 tickets through the dashboard alone: 349 on the web, 224 by voicemail, 111 by email, 20 from the customer portal. Proposals, estimates, exports and the sales book all live in the CRM and the Proposal Engine. Nothing runs on the old sales system any more.

The AWS copy exists and answers in test

Proposal, Services and CRM, with their data, were copied to ANC's AWS account on 26 October and all four test addresses answered today. Production still runs on the current server. The final switch is waiting on one thing: ANC's teams finishing their test pass and giving the word.

Design still lives in the old project tracker to retire

Every design request in the last ten weeks started in Wrike (405 of 420), as did 800 of 810 logged hours and all 6 parts orders. The dashboard mirrors it, and as of this week that mirror is exact and hourly. The team has not moved yet.

Several accounts are still in my name open

The current server, the ancsports.net domain, the AI provider accounts, the usage analytics, the code repositories and the off-site backup storage. They all work. None of them is ANC's yet.

The September plan asked for a decision on the 2027 shape by 14 October. That decision is still open, and it is the one thing on this page that only ANC can close.

02 · Track A

Finish the move to AWS

Done and checked

  • Three separate applications on ANC's panel, each with its own database, worker and supporting services, under Charlie's account.
  • All three databases restored and record counts matched; about 6 GB of backups copied off-host with matching checksums.
  • Sync and rollback rehearsed: new records and files written during the rehearsal survived it. Proposal uploads survive an application restart on AWS.
  • Real work exercised on the copy: an RFP read, edited, saved and reopened; PDF and Word exports; tickets created, edited, reopened, exported; schedule exports matched all 17 sampled events; CRM counts and pipeline totals matched; one labelled Slack message and one labelled email delivered.
  • Scheduled jobs inventoried, including timezone behaviour, and deliberately left switched off on the copy during testing.

Left before the switch

  • ANC's test sign-off. Charlie asked the sales team on 28 October to work their normal CRM day on the copy. The same pass from the service team on the dashboard and the estimators on proposals, then one message: ready. ANC
  • The jobs that make the platforms feel alive. 54 scheduled jobs and 10 background services run on the current server outside the panel: the new-win alerts, ticket digests, technician schedule notices, the email-to-CRM intake every five minutes, the Wrike mirror, the CRM safety guard, the Slack assistant's connection, the nightly backups. They are inventoried and documented; they are not yet built on AWS. I stand them up switched off, and they flip on at the switch while the old ones flip off. Ahmad
  • Re-pointing. The Slack apps, Microsoft sign-in, the email sender and the three domains all point at production today. Each moves in the same hour as the switch, with the list written down in advance. Ahmad · Charlie
  • Backups from AWS. Today's nightly backups run from the current server to storage on my account. Before the switch, the same job runs from AWS into ANC's own storage, and a restore from it is proven once. Ahmad · Charlie
  • Charlie restores production from a backup, with me watching. The September plan's gate: someone other than me has done it. Not done yet. Best done on the AWS copy the week before the switch. Charlie

The switch itself

A weekday evening, Eastern time. About one hour read-only: freeze writes on the current server, run the final data sync we rehearsed on 26 October, move the three domains, turn the jobs on, run the same checks the teams ran in test, open the doors. The current server stays up read-only for fourteen days as the rollback, then goes dark. Proposed: agree the date on 4 November, switch in the week of 17 November, before Thanksgiving week.

03 · Track B

Retire the old tools

  • Wrike, design side. This week the mirror went from nightly to hourly, the tracker's own test folders stopped appearing as requests (222 removed), and 4,718 requests that sat under no client were placed under their venue. What remains: 154 client codes the dashboard has no venue for, 285 open requests on them; Charlie has the list, and each one he confirms fills in on the next hourly run. Then the people change: designers create requests in the dashboard from a set Monday, Wrike goes read-only two weeks later, the mirror is switched off two weeks after that. Charlie · design team · Ahmad
  • Wrike, parts requests and time. Parts requests still go through the Wrike form by Gianni's rule (27 October), and hours are logged there. Both move in the same cutover; the dashboard already holds the parts, inventory and hours records. The request form on the dashboard side is the piece to confirm with Gianni before the date. Gianni · Ahmad
  • Airtable. No record has changed since 9 April. The nightly import of it still runs; I switch it off this week. Nothing else to do. Ahmad
  • The operations tables. Still live: walkthrough logs and open issues, linked to tickets by three jobs every five minutes. They move to AWS with everything else. Whether they are retired into the dashboard is a separate decision for the list Joe is sending; it is not blocking anything. Joe
  • The old sales system. Off since 5 September. The CRM is the only source. —
04 · Track C

What is still not done

The honest list. Each line names the gap, not the excuse.

  • Proposal uploads on production live inside the application itself, not on durable storage. A rebuild of that application loses them. The AWS copy already has this fixed. One more reason not to let the switch drift.
  • The Service Dashboard's database runs as a plain container on the current server, outside the panel's backups and controls. On AWS it is a proper managed service. Same answer: switch.
  • Accounts in my name. Server hosting, the ancsports.net domain registration, the AI provider accounts, usage analytics, the code repositories, off-site backup storage. Each needs an ANC-owned account to move into, and the move is an afternoon once the account exists.
  • No one but me has restored production from a backup. The rehearsal on 26 October was mine. Charlie does one on the AWS copy before the switch; that closes it.
  • Operator handbooks. The developer hub is live since 1 October with 29 guides. The per-platform operator handbook for Charlie is the November deliverable, written as he runs the month.
  • The 2027 shape. Due 14 October by the September plan; still open. The ANC Labs proposal is on the table. It needs an answer by 2 December so that 1 January is clean either way.
05 · After the ground is settled

What comes next

  • Joe's list. The Design side first, as agreed on 29 October; the fuller list arrives this week and is worked in the order it is given, through Joe and Charlie only.
  • Venue spare-parts inventory, inside the dashboard. Charlie's description from 29 October: what each venue holds (spare panels, power supplies, replacement parts), what gets used on a repair and what is left, what ships, whether it actually left and whether it arrived. The dashboard already carries inventory, parts and parts-order records and a low-stock notice; the gap is the per-venue stock and the shipped-received trail. Scoped properly once Joe confirms it belongs on the list, not before.
06

The calendar

Wed 4 Novtouch-base
Teams' test sign-off, or what is blocking it. Switch date agreed. Charlie's client-code list confirmed. The 2027 shape discussed with the Labs proposal in hand.
Mon 10 Nov
Designers create requests in the dashboard from this Monday. Wrike stays open for reading and for anything already in flight.
Week of 10 Nov
Scheduled jobs and background services built on AWS, switched off. Charlie restores production from a backup on the copy. Backups from AWS into ANC storage proven.
Week of 17 Novproposed
The switch: one evening, about an hour read-only. Old server read-only for fourteen days.
Mon 24 Nov
Wrike read-only for design, parts and hours. Everything new starts in the dashboard.
Wed 2 Dectouch-base
Close-out walked line by line: old server off, Wrike mirror off, accounts transferred, handbooks signed off, 2027 answered.
Thu 31 Dec
Term ends. Handover complete either way.
07

What I need from ANC

  • Each team's test pass on the AWS copy finished, and one message saying ready, by 4 November.
  • Charlie's confirmations on the 154 client codes, in any order, as he gets to them.
  • ANC-owned accounts for the six items above, so transfers are not waiting on sign-ups.
  • Gianni's word on the parts request form before 24 November.
  • The 2027 decision by 2 December.

Nothing on this page waits for the 2027 answer. If we continue, this is the ground a lighter engagement stands on. If we do not, ANC ends the year on its own servers, owning every account, with the old tools off and the knowledge written down.