Loading your dashboard…
Our values
- God first — we trust God for wisdom and guidance.
- Real honesty — we tell the truth and communicate clearly.
- Intentional focus — do what matters, cut distraction.
- Total effort — give our best, always improving.
- Consistency builds trust — show up every day and deliver.
What we do & for whom
PurposeHelp families & providers find long-term care with speed, compassion, and clarity — honoring God, empowering caregivers.
Brand promise“Fast, Free, Compassionate Care Placement”
FocusFast, safe long-term care placements using personal service + smart software.
10-year goalA nationally known care platform helping 100,000 patients find safe placements.
3 years out — April 2028
Revenue
$2.28M
Profit
$684K
Clients/mo
150
Referrals/mo
40
This year — by May 24, 2026
Revenue
$1.2M
Clients/mo
75
Facilities/mo
50
Hospital partners
1
Big priorities this quarter (90 days)
1 · Set up our baseline numbersKP
2 · Finish the app — facility login + bed statusKP
3 · Auto referral intake + AI reply under 60sKP
4 · Launch bed-status textingKP
5 · Two-way referral textingKP
6 · Collect data to set baselinesKP
Problems to solve
We work these in the meeting: spot it · talk it · fix it.
- Medicaid billing not filed on time (the big one).
- Some referral data still typed by hand.
- Not always told when caregivers do assessments.
- Need a caregiver questionnaire (auto-text to patient/POA).
- Show a bed-availability map.
Weekly numbers
Checked every week.
| What we track | Owner | Goal | This wk |
|---|
Leadership & hiring plan
Placement runs mostly on automation + AI. Kevin today · key hire · later / PT · Sky/auto
Big-picture leaderKevin
Operations leaderKevin
Office manager — both sides (remote)Maricel
Head nurse — assessments, admissions, 6-mo/annual visitsAga
RN Manager — triage + small caseload + assessments + facility referralskey hire
Field RNs — assessments & caseloadper-action → PT
Finance / moneyKevin
Automation & AI — placement pipelineSky
Departments — our org structure
Loading…
Roles — what each seat owns (Kevin covers most today)
Loading…
Team
Loading…
Team meeting
Now (interim)Sunday — Kevin, Aga, Maricel (time TBD)
TargetTuesday 9:00am once Aga is full-time
LengthsWeekly 60 · Monthly 75 · Quarterly 90
Agenda
Wins & good news5 min
Check the numbers5 min
Check the priorities5 min
News5 min
Task list5 min
Solve problemsmost of it
Wrap up5 min
🎓 Long-Term Care Knowledge
Every JABSOM Care Home ECHO training, watched and summarized by Sky — so the whole team can educate families like experts. Click a card for the full breakdown; drag ⠿ to reorder.
How the work flows
Adult Care Finder places · Case management oversees.
Placement — Adult Care Finder
1
Referral comes in
family · hospital · form
2
Intake & needs
care level · budget · location
3
Nurse assessment
mostly digital · qualify for placement
4
Find matches & tour
real-time beds · visits
5
Placement & bill
move-in · referral fee
Ongoing — Case management
6
Admit to caseload
7
Care plan & oversight
monthly visits · insurance · Medicaid billing
Loading the company scoreboard…
Department KPIs
Referrals / mo
—
Placed / mo
—
Active facilities
—
Reply speed
<60s
Live once tracking + the pipeline run.
Active caseload
~20
Medicaid rate
$14.60/day
Billable days filed
—
Monthly visits
—
Insurance lives here (patient & caregiver). Billing-days is the priority KPI.
Finances
From QuickBooks (Advanced Case Management LLC).
2026 income (YTD)
$15,396
2026 expenses
$20,397
2026 net
−$5,001
Improving: 2025 net was −$34,282; 2026 on pace for a far smaller loss.
Unbilled revenue: ~20 cases × $14.60/day ≈ $8,760/mo possible vs ~$2,566/mo booked. The gap is unfiled Medicaid billing — the #1 fix.
Where the money goes
Payroll (wages + taxes)≈ 99% of expenses
Software, fees, accountingsmall
Cleanup: split income into real streams, merge duplicate wage/tax accounts.
Unit economics
The money math behind the business — your numbers vs. the targets from The Mathematics of Business (Hormozi) and the LTGP:CAC engine. Non-PHI: aggregate averages only, never per-patient.
The 5 numbers to know cold
Green = at/above target · yellow = close · red = a lever to pull. Profit & revenue come from QuickBooks (firm up after Sat cleanup); the rest compute live from your inputs below.
Your inputs — edit to match reality (saved on this device)
Private — you control the rate
Medicaid — state-locked rate
Whole business
Private vs Medicaid — side by side
| Metric | Private | Medicaid | Target |
|---|
Medicaid = $14.60/day × days in month (can't change it, ~45-day pay lag, more paperwork). Private = $600/mo standard ($500 floor), paid at admission.
Hormozi's 12 rules of thumb — your targets
| Price vs close rate | Close >80% = underpriced 3–4× · 30–40% = priced right |
| LTGP : CAC by humans | SaaS 3:1 · +1 human 6:1 · +2 9:1 · all-human (you) 12:1 |
| Gross margin | Services ≥ 80% (net can never beat gross) |
| Payback period | Recoup CAC within 30 days |
| 30-day cash collected | Collect > (delivery cost + CAC) in 30 days |
| Retention | Keep ≥ 80% of cases year-to-year |
| Lead response | Call every new referral < 60 seconds (4× cheaper) |
| Rule of 100 | 100 outreach actions/day × 100 days on one channel |
| Calendar utilization | Keep schedules ~70% full (60–85%) |
| Prepay / cash forward | Discount + bonus → 30–40% prepay |
| Conversion benchmarks | In-person lead 10% · in-home close 35% · web page 1–2% |
| Industry averages | Ignore them — play to win (physics is the only limit) |
The 8 levers to move the ratio
LTGP Raise the private rate
LTGP Cut cost to deliver
LTGP Keep cases longer (retention)
LTGP Cross-sell: ACF placement → ACM case mgmt
LTGP Upsell a higher tier of service
LTGP Downsell — save the “no”
CAC Sharpen the offer to facilities
CAC Fix the funnel — the follow-up engine
Your strategic questions — my read
1-for-1 (private ↔ Medicaid): Smart — it wins the facility as a one-stop shop while the private case carries the margin (~2× Medicaid's, and paid day 0 vs ~day 45). Use private to subsidize thin Medicaid.
⚠️ Only when paired: offer the low-margin Medicaid placement only when a paying private comes with it. If the Medicaid mix climbs too high, blended margin sinks — watch the blend in the 5 numbers above.
All-in private vs. grow Medicaid?
Until QuickBooks is clean, treat Medicaid as volume + relationship + future leverage (prove placement speed → earn paid Medicaid referrals from Kaiser/Straub), not a margin engine. Scale private; use Medicaid to win the relationship.
Referral reward “horse race”
Charity-to-the-whole-office (not an individual) is the right call — it sidesteps the kickback / anti-inducement problem. Track referrals by facility (a leaderboard) — that's your CAC attribution, and the next thing worth building here.
CEO command center
▸ Daily brief — your 7 AM plan + 5 PM wrap, tap to open
The 5 numbers — know these cold
Department KPIs — tap to drill in
▾ Agentic OS — departments → agents → skills → automations
Loading…
Process map — referral → 6-month visits — tap a stage to see what's inside Open team process maps
Automation health — live monitor, auto-checked every 3 hrs
Loading…
Inbox — Emily — your email secretary: instant watch alerts + inbox triage
Loading…
Follow-up engine — open GHL pipeline cards; red = sitting 3+ days (the dropped-follow-up bottleneck)
Loading…
Systems & spend — every tool, what it costs, and how close it is to its limit
Loading…
Money
Cash & runway
pending cleanup (Sat)
Unbilled / leak
~$6K/mo
Owed to you (A/R)
$628
Savings (clean)
$15.9K
Bank balances read wrong (unreconciled) — real cash & runway after Saturday's QuickBooks cleanup.
Needs your decision
Approve the QuickBooks cleanup planSat
Aga full-time vs. hire a clinical manager (confirm model inputs)soon
Pick the Sunday meeting timethis wk
This week — top 3
1 · Clean QuickBooks → unlock the real numbersSat
2 · Get Aga + Maricel using the dashboardnow
3 · Clear the Medicaid billing backlogleak
Pulse
Scorecard
0 of 6 filled
Quarterly rocks
6 · set status
This week's meeting
Sun (interim)
Critical fires non-PHI
Medicaid billing not filed — revenue leakKP · Sat
Vendor BAAs (Anthropic / Airtable)KP · Sun
Wins this week
- Dashboard live at dashboard.acmahi.com
- GoDaddy + Resend connected via API
- Team accounts created (you, Aga, Maricel)
Key dates coming up
QuickBooks cleanupSat Jun 27
Vendor BAAs dueSun Jun 28
Aga RN-workflow interviewSun Jun 28
AI capability briefing — what today's AI can actually do, from The Rundown AI (free · 2M readers)non-PHI · updated 07/11/2026
Not healthcare-specific — this is your "what's possible" radar. Plain-English read of the frontier, and which parts touch ACM.
Loading latest issue…
🔎 What AI can do right now (that matters to us)
Workplace agents (ChatGPT Work, Claude Cowork) — AI that does multi-step work on your computer: browses, writes code, fills forms. Not just chat. → This is exactly the Sky/Code path we're already building.
AI that audits itself (Claude "Reflections") — reviews how you use it and suggests better workflows + skills. → Fits our AI-spend & automation tracking.
Cheaper big-brain models (Meta Muse Spark: 1M-token memory at ~¼ the price) — more automation for less money.
Token-saving orchestration — use one smart model as planner + cheap models as workers → cut AI cost ~60%.
✅ Could we use it?
Yes — as a radar, not an ops tool. The newsletter is general AI news (low healthcare signal), so treat it as 5-min/day learning. But two items map straight onto ACM: Claude Reflections (try now) and the planner/worker cost pattern (how Code/Sky should already run). Don't wire any automation to the newsletter itself — it's a learning feed, not a system of record.
🎯 Action steps
1 · Bookmark the free archive; skim 5 min/day; funnel keepers into the vaulteasy
2 · Try Claude "Reflections" to audit our Claude usage + get skill ideasthis wk
3 · Apply planner/worker orchestration to cut Code/Sky token spendL2 fit
4 · Keep ops automations OFF it — general AI news, not a record systemguardrail
Roadmaps
Focus: two builds unlock everything above — the follow-up engine (L2) and clean QuickBooks (L4). Finish those two before anything else.
A&K Update
Today
Win to celebrate today
Today — what I worked on / finished
Need Aga's eyes
Shared to-do
Tap the status to cycle: Scheduled → In progress → Pending → Done.
Reports
What got done — and whether it was on track. Auto-logged as we work together.
Newsletter
Newsletters from CTA, DOH & partners — the original, a plain-English summary, how it affects us, and action steps. Auto-collected; review and clear.
Loading newsletters…
Company Hub
The company's go-to reference — subscriptions, logins, SOPs, and who to call. Replaces the Airtable Operations HQ.
Tip: drag a tab to reorder it — your layout is remembered on this device.
🌱 Personal Tracker
Habits, goals, daily check-ins and the people you invest in. Just for you — no one else can see this tab or its data.
📥 TCC Library
Dump video links here → I transcribe each one and report back: what it's about, what's useful for the agency, and tasks to approve. Just for you · processed on your PC.
Drop links
Paste as many as you want — one per line. They go into the queue and get processed one at a time.
⏳ Queue
✅ Processed
📚 Records — kept for reference; 📌 pinned = build next
🔑 Prompt Key
Your shortcut triggers across every tool — what to type and what it does. Just for you · saved on this device.
Mentors I model
People whose wisdom, ethics and walk I want to mirror. When I send a TCC video from anyone here, weight the takeaways toward following their lead — and tell me what to copy.
Faith
⭐ Wes Huff apologetics
⭐ GodLogic Apologetics apologetics
⭐ Bryce Crawford evangelism
⭐ Nicholas Bowling evangelism
⭐ Nick Freitas + politics
⭐ Charlie Kirk + culture
⭐ Daniel Adams supernatural · TSNL
Business
⭐ Myron Golden + faith
⭐ Ryan Pineda + faith + family
⭐ Patrick Bet-David + faith
⭐ Taylor Welch + faith
⭐ Dan Martell + faith
⭐ Ryan Blair + faith
Alex & Leila Hormozi
Codie Sanchez
Ken McElroy real estate
Pace Morby real estate
Sharran Srivatsaa real estate
Mark Moss finance
Neil McCoy-Ward finance
Ryan Deiss marketing
Jeremy Miner sales
Sandeep Swadia AI · MIT Monk
Simon Sinek leadership
Tom Bilyeu
David Meltzer
Jesse Itzler
Marques Brownlee tech
Ron White memory
Health
⭐ Bobby Parrish + faith · food
⭐ Dr. John Delony + faith · mind
Andrew Huberman neuroscience
Bryan Johnson longevity
Paul Saladino nutrition
Dr. Eric Berg nutrition
Parenting & family
⭐ Scott Donnell + faith + business · kids & money
Family wisdom also flows from the starred Ryan Pineda & Dr. John Delony above.
News & culture
Promethean Updates geopolitics
Web-check notes
Checked, no clear public profession of Christ → in Business unstarred: Jeremy Miner · Sharran Srivatsaa · Neil McCoy-Ward. Say the word if you know otherwise.
Hale Ledger — receipts & expenses by property (Dan & Rochelle)
Open full tab
Growth Hub — family learning & mini-courses
Open full tab
Guitar Coach — learn guitar in the key that fits your voice
Open full tab
Camp Planner — trip planning
💪 Stronger — Kevin & Aga's kettlebell workout timer
Open full tab
🧊 Freezer — what's in the freezer, what to use first, what you can cook tonight
Loading…
ACF Portals — the three LIVE portals · open here or scan with your phone
🧑⚕️ Caregiver portal
Caregivers sign in by text, browse de-identified referrals, schedule assessments, and manage their home profile (photos, bio, video tours).
👪 Family page
Each family gets a private texted link (no login): compare interested homes, schedule visits, approve. There is no public URL — the demo below uses TEST data.
📋 Referral partner portal
Social workers & case managers sign in by text and see the live stage of every referral they sent — plus pre-admission fields they can fill in directly.
🩺 Care-Level Estimator STAFF
Your phone-triage tool. Plug in a caller's answers → estimates their DHS 1147 functional points vs the 15-pt foster-home gate → tells you: book the RN assessment, or refer out. Internal only — not public, not an official 1147.
📄 1147 Generator STAFF
Search a patient, click one button — stamps the P2 clinical answers already in GHL onto the real DHS 1147 PDF (sections III–XIII circled, points totaled) so you're not re-typing the assessment by hand. Saves to S3 and downloads.
🗺️ Process Maps TEAM
Referral to Placement + Case Management — the whole business as a living map the TEAM edits. Add cards for the tasks YOU do, tag who owns them + what's automated, list the forms, upload blank + John-Doe examples, attach old SOPs/Scribes. Red = files still needed. First pass due Mon 07/13.
🛏️ Availability Map STAFF
LIVE map of all ~1,124 licensed foster caregivers (monthly CTA list) at their real addresses — open beds by area with Medicaid / Private / Hospice tabs, plus a toggle to the state's 451 ARCH care homes and 284 Expanded ARCH with vacancies. Use it on calls: family names an area, you see how many beds are open there. Caregivers update bed status in their portal; snapshot rebuilds daily 5:30 AM. Hawaii publishes no foster-home availability — this view exists nowhere else. (Mock-availability demo: add ?demo=1.)
🩺 Field RN Portal STAFF
The field nurse's visit-documentation portal. RNs sign in by text, see the month's assigned visits, and complete the monthly visit note — all eleven HAR §11-800-27(b) monitoring elements — on a phone in the home. Members appear as de-identified codes only. Voice dictation isn't in v1; narrative is typed, pending the recording-consent question.
📱 Scan any QR with your phone camera to open that portal on mobile — same live data as desktop. (Old sample-data prototype: acf-portal-prototype.html)
Loading Sky…
Loading Marcus…
Loading Emily…
🗺️ Business Map — How Our Business Works
The big picture in one look: where referrals come from, the stages they move through, and exactly where the money comes in. Just for you · big picture only, no patient detail. Updated 07/16/2026.
🔵 Adult Care Finder (ACF)
Our public brand. Finds the right foster home for the patient — from the referral coming in through the day they move in (admission).
🟢 Advanced Case Management (ACM)
Our agency. Takes over after the move-in — handles the state/insurance paperwork and keeps managing the care month after month.
💰 Money badge
A gold badge on a step means money comes in at that step — it says who pays and which company gets paid.
1 · 📥 Referral comes in
Hospitals (Kaiser, Straub), home-health agencies, social workers, or families reach out — by fax, phone, or the web — asking us to find care for someone.
➡️
2 · 🧾 Financial check
How will care be paid for — Medicaid or private pay? We verify eligibility and chase any missing documents.
➡️
3 · 🩺 Nurse assessment
Our nurse reviews the patient's needs to figure out the level of care the home must provide.
⬇️
4 · 🏠 Find caregivers
We search our network of licensed foster homes and contact caregivers who can take this patient.
⬅️
5 · 🤝 Family tours & approves
The patient / family (POA) visits the interested home(s) and picks the one they like.
⬅️
6 · 📄 Final documents
Home approved — we gather the last paperwork needed to complete the move-in.
⬇️
7 · 🎉 Admission (move-in day)
The patient moves into the foster home. ACF's job is done here.
💰 Caregiver pays ACF — 50% of the home's first month 💰 Private pay: family pays ACM $600/mo — starts today➡️
8 · 🗂️ State & insurance paperwork
ACM files the backend paperwork with the state and insurance (especially Medicaid) so ongoing care is covered.
💰 Medicaid: state pays ACM ~$438/mo — arrives ~45 days later➡️
9 · 🔁 Follow-up visit & every month after
About a month after move-in, our nurse visits the home — and keeps visiting monthly while we manage the case.
💰 Monthly case-management fee keeps coming in to ACM💰 Where the money comes in — the whole story in 3 lines
At move-in (step 7)The caregiver (foster home) pays Adult Care Finder a one-time fee: 50% of the home's first-month total. This is the placement fee.
At move-in (step 7), private payThe family/POA starts paying Advanced Case Management $600/month for case management — from day one.
~45 days later (step 8), MedicaidThe state (Medicaid) pays Advanced Case Management ~$438/month (~$14.60/day) — slower to arrive, then monthly.
Two companies, two customers: the caregiver pays ACF once for the placement; the family or the state pays ACM every month for ongoing case management.
📊 The Rate Fight — Program History in Pictures
How Hawaii's foster-home program was built, what got piled on us over the years, and the one chart that says it all: caregivers got raises, agencies got frozen. Just for you · every number web-verified · sources in the vault. Updated 07/17/2026.
🏠 CAREGIVERS (foster homes)
+43%
Medicaid day-rate rose from $47.06 (2020) to $67.33 (2026) — including a +32% jump in 2025 from HB2215 funding.
🗂️ AGENCIES (case management — us)
+11%
Rate crept from ~$13.15/day (2021) to $14.60 — then FROZEN for 2026. That's about +1.3%/yr over 5 years, well under inflation.
💵 The one chart that says it all — same program, two very different raises
💡 The state's own actuary (Milliman) said $15.06–$16.48/day was fair for agencies. They pay $14.60 and froze it. Fixing it statewide costs only $1.3–2.3M/yr — the cheapest item in the whole study.
📜 How the program was built — the 47-year story
1979
🏥 The idea starts
Queen's Medical Center pilots foster homes for nursing-home-level patients — family-home care instead of an institution.
1984
💰 Medicaid starts paying
A federal waiver funds it — approved because foster homes cost the state two-thirds less than a nursing home. That savings is our whole argument.
2001
📖 It becomes law
Act 273 makes it official and requires every foster-home client to have a licensed case management agency — that requirement is why we exist.
2002
📋 First rulebook (HAR 17-1454)
The first licensing rules for agencies and homes: RN case managers, monthly visits, 6-month reassessments — the core duties we still do today.
2018
🔄 New rulebook, same duties (HAR 11-800)
Oversight moves to the Dept. of Health. The rules were copied over nearly word-for-word — proof our core job hasn't shrunk since 2002.
2020
⚠️ Fines arrive
The state starts fining agencies AND homes — up to $500 per violation, per day. New risk, no new pay.
2025
📈 Caregivers get +32%
HB2215 funds a big caregiver raise. Case management moves $14.44 → $14.60 (+1.1%).
2026
🧊 Agencies frozen
QI-2604 holds case management at $14.60 while a new 10-day / $500-a-day paperwork rule lands. The rate discussion is open at MQD right now.
🧱 More piled on us every year — while the pay line stayed flat
$13
2018
HAR 11-800
$13
2020
+ fines regime
$13
2022
+ in-person only
$14.44
2024
+ background checks + HFA
$14.60
2026
+ HOKU + 10-day POC
🧱 = a new requirement layered on (fines, background-check architecture, in-person-only visits, HOKU enrollment, HFA home-safety checks, 10-day correction deadlines). The green number underneath is the daily pay — nearly flat the whole time.
🎯 The takeaway (bring this to the coalition)
- Caregivers +43%, agencies +11% — same program, and the mandatory oversight layer got skipped.
- The state's own numbers back us: Milliman said $15.06–$16.48/day; MQD pays $14.60 and admitted (2021) the rate was flat for a decade.
- It's cheap to fix: $1.3–2.3M/yr statewide — the smallest line in the entire rate study.
- Duties only grew: unchanged in the rulebook since 2002, but fines, background checks, HOKU, HFA, and 10-day/$500-a-day deadlines were stacked on top.
- The window is now: the Chapter 800 rewrite is in draft and MQD's rate discussion is open — before any public notice.
Full sources: Wiki — CCFFH/CMA Program History & Rates · Compliance — CTA Requirement-Creep Timeline · Compliance — CMA Duty Inventory (all in the vault).
📱 Foldable Phone Tracker
Two foldables — one for you, one for your wife · Verizon (Hawaii) · trade in Note 10 + Note 10+. Just for you · auto-updates every Wednesday. Updated 08/26/2026.
✏️ This week's edition was rewritten after an audit returned FIX
An adversarial audit of my first draft found 21 defects, six of them user-facing — including a false "reconciles exactly" claim, a combined trade-in range that didn't match its own components (the same defect I publicly corrected last week), Samsung credit figures presented as Razr Fold prices, and a recommendation that contradicted my own stated belief. Corrections are marked ✏️ AUDIT FIX below. The recommendation changed as a result — see the call.
🔴 The big Verizon credit may need a NEW LINE — which you may not have
Android Central's Verizon deals guide (updated 08/20) spells out the terms for both foldables, and it isn't two conditions — it's three, stacked: "$1,100 off by trading in and adding a line with Unlimited Ultimate" (Z Fold 8), and "up to $1,100 off when you process a trade-in and add a line with the Unlimited Ultimate plan" (Fold 8 Ultra). Every other free-phone offer on that page — Pixel 11, Galaxy S26, Razr 2026, Edge 2026 — carries the same "add a line" wording. Why it matters to you: if you and your wife are upgrading two lines you already have rather than adding two new ones, the ~$1,100-per-phone credit may not be available to you at all — and the prize I've been telling you to wait for could be zero.
⚠️ How strong is this? Weaker than I first wrote it. (a) The source is Android Central, a secondary site that earns affiliate commission on these links — specific and internally consistent, but not Verizon. (b) I could not read the $1,100 terms on Verizon's own site — their deals pages 404'd again, second straight week. (c) ✏️ AUDIT FIX — I've withdrawn my third piece of support entirely. I'd cited clicking the "Existing customer" toggle on Verizon's Z Fold 8 page. It tested the wrong offer — the toggle sat on the $460 Simplicity credit, which by its own terms needs no new line and no trade-in, so it could say nothing about the $1,100 question. Recorded as NA, no evidentiary weight. (d) ✏️ AUDIT FIX — counter-evidence I shouldn't have dropped. Last week's edition carried a Verizon primary source on this exact offer: their own 07/22 release prices the Fold 8 Ultra at $24.99/mo × 36 on Unlimited Ultimate = $899.64, i.e. $1,200.35 off — and it names no new-line condition. My first draft silently deleted it and let a weaker secondary source become the headline. Both now stand side by side, and I'm not going to pretend they resolve: either the press release omitted qualifying conditions (they routinely do), or the terms tightened between 07/22 and 08/20. I don't know which.
⚠️ How strong is this? Weaker than I first wrote it. (a) The source is Android Central, a secondary site that earns affiliate commission on these links — specific and internally consistent, but not Verizon. (b) I could not read the $1,100 terms on Verizon's own site — their deals pages 404'd again, second straight week. (c) ✏️ AUDIT FIX — I've withdrawn my third piece of support entirely. I'd cited clicking the "Existing customer" toggle on Verizon's Z Fold 8 page. It tested the wrong offer — the toggle sat on the $460 Simplicity credit, which by its own terms needs no new line and no trade-in, so it could say nothing about the $1,100 question. Recorded as NA, no evidentiary weight. (d) ✏️ AUDIT FIX — counter-evidence I shouldn't have dropped. Last week's edition carried a Verizon primary source on this exact offer: their own 07/22 release prices the Fold 8 Ultra at $24.99/mo × 36 on Unlimited Ultimate = $899.64, i.e. $1,200.35 off — and it names no new-line condition. My first draft silently deleted it and let a weaker secondary source become the headline. Both now stand side by side, and I'm not going to pretend they resolve: either the press release omitted qualifying conditions (they routinely do), or the terms tightened between 07/22 and 08/20. I don't know which.
🟡 ✏️ AUDIT FIX — my "$460 reconciles exactly" claim was wrong, and this one stings
I published that last week's $20 discrepancy was "resolved," that the numbers "reconcile exactly," and marked it "Closed." That's false. Verizon shows a $460 credit and a $39.99/mo price, and they don't agree: $1,439.99 ÷ 36 = $39.9997, which rounds to $40.00, not $39.99. Inverted: 36 × $39.99 = $1,439.64, implying $460.35 off. A third route — 36 × ($52.77 − $39.99) = $460.08 — gives yet another value. What actually happened is the gap shrank from $20.35 to $0.35. It did not close. Immaterial to your wallet; material to whether you can trust me when I write "exactly" — I used it for three numbers that don't match.
🟡 I over-read the used market last week, and this week it reversed
Last week I reported supply tightening (14 → 12 listings) with prices ticking up, and used it to walk back my earlier "waiting is safe" claim. One more week of data went the other way: 15 listings now (up from 12), the floor dropped $1,082 → $1,049, and completed sales dropped $1,133 → $1,110. Two weeks of opposite-direction noise is the signal: the used market is roughly flat — not reliably falling, not tightening. I should have said "flat" a fortnight ago instead of narrating each week's wobble as a trend.
The pick
Razr Fold
Two new, all-in
$3,999.96
Two used, all-in
~$2,746
This week's call
Conditional · likely BUY USED
✏️ AUDIT FIX — the recommendation itself changed this week
My first draft said WATCH flatly while admitting, in the same paragraph, that I suspect you're upgrading existing lines — the very condition under which my own stated flip rule says the answer is BUY USED. The audit was right that this is incoherent: I was recommending against my own best estimate and making the flip contingent on you confirming something I already believed. So, stated properly:
CONDITIONAL. If you're ADDING two new lines → WATCH, hold to the 09/30/2026 cutoff. If you're UPGRADING two existing lines → BUY USED now. My own best guess is that you're upgrading, which means my actual recommendation today is BUY USED — and I'll hold that unless you tell me otherwise.
Why the conditional is the honest form: the entire case for waiting is a Verizon credit that may require a new line. I can't confirm the requirement (secondary source only) and I can't confirm your line situation. But I shouldn't present the wait as unconditional when I privately believe the condition fails. If you are waiting, the cutoff holds: if Verizon hasn't listed the Razr Fold by 09/30, buy two USED (~$2,746) rather than new ($3,999.96).
CONDITIONAL. If you're ADDING two new lines → WATCH, hold to the 09/30/2026 cutoff. If you're UPGRADING two existing lines → BUY USED now. My own best guess is that you're upgrading, which means my actual recommendation today is BUY USED — and I'll hold that unless you tell me otherwise.
Why the conditional is the honest form: the entire case for waiting is a Verizon credit that may require a new line. I can't confirm the requirement (secondary source only) and I can't confirm your line situation. But I shouldn't present the wait as unconditional when I privately believe the condition fails. If you are waiting, the cutoff holds: if Verizon hasn't listed the Razr Fold by 09/30, buy two USED (~$2,746) rather than new ($3,999.96).
🥇 Motorola Razr Fold The pick · settled
Verizon status⏳ Still not launched — 97 days (just under 14 weeks) after the 05/21 unlocked launch. ✏️ AUDIT FIX: I'd written a flat "14 weeks"; 05/21 → 08/26 is 13 weeks 6 days. I read Verizon's Motorola page live today: 17 phones listed, no Fold. The count held at 17 this week (it fell 18 → 17 last week), so the page stopped shrinking. The obvious direct URL, verizon.com/smartphones/motorola-razr-fold-2026/, returns "this page no longer exists."
✅ Works on Verizon?✏️ AUDIT FIX — now sourced, not asserted. I'd been claiming "the unlocked model works fine on your Verizon lines" for weeks with nothing behind it, while two rows below ruling OUT the OPPO on a full band enumeration — an asymmetric standard on the one assumption the entire BUY USED route depends on. Checked today: Best Buy's listing for the unlocked razr fold 2026 (PBAY0003US) names Verizon explicitly among compatible carriers, alongside AT&T, T-Mobile, Metro, Cricket, Google Fi, Visible and Mint. The route is sound — but because I checked, not because I kept saying so.
Launch signal⚪ No movement, fourth week running. The Motorola "razr fold – 2026 (Verizon)" support page is in exactly the same state — guides posted, videos still "coming soon." Being honest: that hint hasn't gone anywhere in a month.
Best current deal$1,899.99 (512GB) direct — unchanged for the fourth straight week. ✏️ AUDIT FIX: I'd called this "byte-for-byte unchanged." I never made a byte comparison — I read the text verbatim each week and it read the same. Today, verbatim: free Moto Tag + Moto Watch + Moto Buds Loop ($429.97 value, while supplies last), pen still not included, plus a $300 trade bonus you can't use here.
Trade-in🚩 Not available in Hawaii — the exclusion is printed on the Razr Fold's own checkout page: "Trade-in program not available for orders shipping to Alaska, Hawaii, or Puerto Rico." Verizon is your only trade route, and it still doesn't sell this phone.
⚠️ Don't buy the wrong oneDo NOT confuse this with the "motorola razr 2026" Verizon sells at $799.99 — that's the FLIP Razr, a different and much cheaper phone. Verizon gives that one away free on Unlimited Ultimate, which is exactly why it's easy to click by mistake. You want the book-style Razr Fold, 8.1" inner screen, $1,899.99.
Stylus✅ Supported — Moto Pen Ultra, verified live today at $99.99, no promotion on it. Two pens = $199.98. Still the only book-style foldable with a working stylus you can run on Verizon — this is the whole reason the phone wins.
Used marketSwappa today: 11 unlocked listings, avg ask $1,273 — supply up from 8, price flat (was $1,275). Also 4 T-Mobile-locked at ~$1,219 (that pool got dearer, and it's wrong for Verizon anyway). Zero Verizon-branded, as expected. Pool-wide: floor ask $1,049, avg ask $1,258, avg completed sale $1,110.
Two used, all-in~$2,746 at the asking average (2 × $1,273 + $199.98 pens), or ~$2,420 if you transact nearer completed-sale prices (2 × $1,110 + pens). ✏️ AUDIT FIX — why asks here but sales for your Notes below, which I failed to explain: the audit caught that both choices happened to push the same way (making waiting look better), and that's a fair hit. But the fix is to state the reason, not change the number — as a buyer you pay close to the ask, because you buy from a live listing; as a seller you receive close to the completed sale. Asks for buying, sales for selling is the correct asymmetry. Both figures are now shown in both directions so you can check me.
📱 Samsung Galaxy Z Fold 8 / Fold 8 Ultra ❌ Ruled out · no S Pen
Stylus (S Pen)❌ RULED OUT — confirmed, re-verified today, no change. Neither the Fold 8 nor the Fold 8 Ultra supports the S Pen (no digitizer — dropped with the Fold 7 for thinness, never restored). Buying an S Pen separately doesn't help; it won't pair. That's your deal-breaker → Samsung stays out.
🔎 Worth knowingSamsung EVP HS Moon says they're "exploring" bringing the S Pen back on a future Fold. Nothing committed — no model, no date. Samsung isn't permanently out of your consideration, just out for this generation. Revisit around Fold 9 (mid-2027).
Launch / priceOn sale now (retail from 08/07), though Verizon's page today shows shipping 08/28 – 09/11 — roughly a two-week backorder. Z Fold 8 = $1,899.99; Fold 8 Ultra = $2,099.99.
Verizon dealThree published figures, and they don't all agree: (1) $460 device savings on the Z Fold 8, Simplicity Plan required, no trade needed — Verizon's own page, but see the $0.35 problem above. (2) Verizon's own 07/22 release: Fold 8 Ultra at $24.99/mo × 36 on Unlimited Ultimate = $899.64, i.e. $1,200.35 off, with no new-line condition stated — and this one does reconcile exactly. (3) Android Central 08/20: up to $1,100 off either foldable, requiring trade-in and a new line and Unlimited Ultimate. (2) and (3) conflict; I'm leaving that unresolved rather than picking the one I like. Plus 6 months Disney+/Hulu/ESPN+ free.
Buy now or wait?N/A — out of the running. Tracked only because it's the only published evidence of what Verizon actually pays out on a foldable.
❌ OPPO Find N6 Ruled out · no change
VerizonNot compatible — supports T-Mobile/AT&T bands (B2/B4/B5/B12/B25/B41/B66/B71) but not Verizon's Band 13 / CDMA-VoLTE fallback. Won't work properly even imported.
U.S. availabilityRe-checked today by search: no U.S. carrier partnerships with Verizon, AT&T or T-Mobile. It's a China-market model; the only U.S. route is grey-market import through resellers. ✏️ AUDIT FIX: last draft said "re-verified" with no URL attached — indistinguishable from carrying the claim forward. Sources are now cited in the note. Static for five months; this is a formality, not a live option.
🔁 Trade-in values — Note 10 & Note 10+
| Your phone | Verizon (base) | Samsung.com | Motorola.com | Sell privately |
|---|---|---|---|---|
| Galaxy Note 10 | Unknown — need IMEI | Not accepted ($0) | N/A in Hawaii | ~$150–$179 · central est. $170 |
| Galaxy Note 10+ | $30.00 (07/16) | Not accepted ($0) | N/A in Hawaii | ~$170–$230 · central est. $202 |
Combined private sale: ~$320–$409 ($150 + $170 = $320; $179 + $230 = $409). ✏️ AUDIT FIX: my first draft printed "$320–$410," which doesn't equal its own components. A one-dollar error — and completely damning in context, because it's the exact defect I publicly corrected last week after it had sat in this tracker for two weeks, and I reintroduced it in the very edition that took credit for catching it. It's now tied to its parts, and I'll check it against them every week.
✏️ AUDIT FIX — the derivations, shown rather than asserted. The audit caught me publishing bands that appear nowhere in the data. Note 10: one Verizon 256GB listing @ $179; pool floor ask $150; pool avg completed sale $170 → range $150–$179, central $170. Note 10+: ⚠️ third straight week with zero Verizon 256GB listings, so this stays an interpolation — start from the 256GB average completed sale of $202, apply ±15% for the carrier-locked discount → $172–$232, rounded to ~$170–$230. For contrast, unlocked 256GB asks actually rose this week (10 listings @ $274, up from 5 @ $266). Note: there is one Verizon Note 10+ listing at $278, but it's a 512GB unit — wrong storage, don't anchor on it.
✏️ AUDIT FIX — the "5–7× better" claim was built on an invented number. I'd been dividing by "~$60 of appraised trade value." That $60 was $30 × 2 — I doubled a figure verified for the Note 10+ alone and applied it to a Note 10 whose Verizon value is explicitly unknown and IMEI-blocked. That invented denominator was live for two weeks. What I can honestly say: on the one phone where I have both numbers, the Note 10+ sells privately for ~$170–$230 against Verizon's verified $30 — roughly 6–8× better. The pair-wise multiple can't be computed until the Note 10 is quoted.
⚠️ Still no live Verizon quote — six weeks blocked. Verizon's estimator won't return a dollar figure without a full 15-digit IMEI, and only a partial one is on file (Note 10+, ending 2678). I re-swept KP Brain, ACM Brain and memory again today — the only "IMEI" hit outside this tracker is a note about how to ask you for them. Dial *#06# on both phones and send me the two numbers.
The crux. For six weeks the theory has been: keep the Notes as trade ammunition, because the Verizon credit dwarfs their cash value. That rests entirely on the credit being reachable. If you're upgrading existing lines, the ammunition may be worth roughly its appraised value (a verified $30 plus an unknown) rather than four figures — in which case selling for ~$320–$409 cash is clearly better, and the ammunition theory collapses.
❓ Three questions, in order of how much money they move: (a) 🆕 are you ADDING two new Verizon lines, or UPGRADING two you already have? — biggest single variable, and it may flip the call today; (b) which Verizon plan are you on? (c) what phones are you and your wife carrying right now — if they're Galaxy S/Z/Note, they become the trade-ins and the Notes are free to sell today.
✏️ AUDIT FIX — the derivations, shown rather than asserted. The audit caught me publishing bands that appear nowhere in the data. Note 10: one Verizon 256GB listing @ $179; pool floor ask $150; pool avg completed sale $170 → range $150–$179, central $170. Note 10+: ⚠️ third straight week with zero Verizon 256GB listings, so this stays an interpolation — start from the 256GB average completed sale of $202, apply ±15% for the carrier-locked discount → $172–$232, rounded to ~$170–$230. For contrast, unlocked 256GB asks actually rose this week (10 listings @ $274, up from 5 @ $266). Note: there is one Verizon Note 10+ listing at $278, but it's a 512GB unit — wrong storage, don't anchor on it.
✏️ AUDIT FIX — the "5–7× better" claim was built on an invented number. I'd been dividing by "~$60 of appraised trade value." That $60 was $30 × 2 — I doubled a figure verified for the Note 10+ alone and applied it to a Note 10 whose Verizon value is explicitly unknown and IMEI-blocked. That invented denominator was live for two weeks. What I can honestly say: on the one phone where I have both numbers, the Note 10+ sells privately for ~$170–$230 against Verizon's verified $30 — roughly 6–8× better. The pair-wise multiple can't be computed until the Note 10 is quoted.
⚠️ Still no live Verizon quote — six weeks blocked. Verizon's estimator won't return a dollar figure without a full 15-digit IMEI, and only a partial one is on file (Note 10+, ending 2678). I re-swept KP Brain, ACM Brain and memory again today — the only "IMEI" hit outside this tracker is a note about how to ask you for them. Dial *#06# on both phones and send me the two numbers.
The crux. For six weeks the theory has been: keep the Notes as trade ammunition, because the Verizon credit dwarfs their cash value. That rests entirely on the credit being reachable. If you're upgrading existing lines, the ammunition may be worth roughly its appraised value (a verified $30 plus an unknown) rather than four figures — in which case selling for ~$320–$409 cash is clearly better, and the ammunition theory collapses.
❓ Three questions, in order of how much money they move: (a) 🆕 are you ADDING two new Verizon lines, or UPGRADING two you already have? — biggest single variable, and it may flip the call today; (b) which Verizon plan are you on? (c) what phones are you and your wife carrying right now — if they're Galaxy S/Z/Note, they become the trade-ins and the Notes are free to sell today.
⏰ Deals alert Watch: Verizon Razr Fold
🆕🔴 New-line claim 08/26The ~$1,100 Verizon foldable credit reportedly requires trade-in + new line + Unlimited Ultimate (Android Central, updated 08/20). Secondary source, not confirmed on Verizon's own site, and in tension with Verizon's own 07/22 release, which stated no such condition. ✏️ AUDIT FIX: this row was labelled "New-line rule" — asserting as settled, in the most scannable spot on the card, what the body correctly hedges. Corrected to "claim."
🆕 Corrected 08/26Verizon's printed Z Fold 8 device saving is $460 (was $480 last week). The gap between the printed credit and the monthly price narrowed from $20.35 to $0.35 but did not close — see the correction card above.
UnchangedMotorola: free Moto Tag + Moto Watch + Moto Buds Loop ($429.97), no pen, plus a $300 trade bonus unusable from Hawaii. Same text for four straight weeks — the standing offer, no end date.
Live now (wrong phone)Verizon Z Fold 8 / Fold 8 Ultra: $460 off on Simplicity (no trade), up to ~$1,100–$1,200 off on Unlimited Ultimate, + 6 months Disney+/Hulu/ESPN+ free. No S Pen, so not for you.
Not actionableT-Mobile still offers the Razr Fold free on Experience Beyond / Go5G Next. Noted only to be complete — you must stay on Verizon, so this isn't a route.
Next trigger🎯 Verizon listing the Razr Fold with a deal — that's what would make your Notes worth something. 97 days and counting. I'll flag it the moment it appears and re-quote both Notes that same day.
Everything elseStill no BOGO or multi-line promo on any current-generation foldable — searched again today, nothing. The 08/12 Flip7/Fold7 BOGO stays parked as unverified (Verizon's deals pages 404'd for the second straight week). None of this week's four standing triggers fired.
💰 The three routes, priced honestly
WAIT for VerizonCredit is UNKNOWN. ✏️ AUDIT FIX — this is the correction that matters most. I've been quoting "~$2,200 across two lines" and "~$920" as if they were Razr Fold numbers. They are not. They're 2 × $1,100 and 2 × $460 — both Samsung Z Fold 8 figures, for a device this same page rules out — used as a proxy for what a Razr Fold deal might look like if one ever appears. The proxy may be reasonable; it is not a price, and I presented it as one ("priced out today"). Honest version: if Verizon lists the Razr Fold, and the promo follows the Samsung pattern, and you meet the conditions, the credit might land somewhere between ~$920 and ~$2,200 across two phones. If Verizon never lists it — the observed state for 97 days, no launch signal in four weeks — the credit is exactly zero, not "near zero."
BUY USED today~$2,746 for two unlocked at the asking average plus two pens, or ~$2,420 nearer completed-sale prices. Available immediately, no Verizon dependency, and Verizon compatibility is now confirmed rather than assumed. Costs the warranty and the gift bundle, and carries hinge-condition risk — the part of a foldable most likely to fail and the hardest to judge from a listing photo.
BUY NEW today$3,999.96 all-in for two phones and two pens, forfeiting any future credit. Gets the gift bundle (Watch/Buds/Tag ×2) — I still haven't verified those gifts ship to Hawaii, and given the trade-in program explicitly excludes Hawaii, that's worth checking before anyone counts on it.
✏️ AUDIT FIXI'd listed "$199.98 of pen forfeited" as a cost of waiting. It isn't — the pens cost $199.98 on every route and are already inside both the $2,746 and the $3,999.96 totals. A sunk cost common to all three; attributing it to one route distorted the comparison.
Full write-up: KP Brain ▸ Resources ▸ Purchases ▸ Foldable Phone Tracker. Flags I'll raise the moment they happen: a Verizon Razr Fold deal, any two-phone foldable promo, an OPPO US/Verizon move, or a big trade-in value drop.
🖥️ Always-Home Machine Tracker
One Windows mini PC, on 24/7, running your Code projects + scheduled automations + Sky assistant, so the Acer stops competing. One status monitor on it · share your laptop's keyboard/mouse (no second set). Just for you · I keep this current. Updated 07/15/2026.
✅ PURCHASED 07/15 — GEEKOM A9 Max · arriving Sun 07/19
GEEKOM A9 Max (Ryzen AI 9 HX 370, 32GB DDR5 upgradeable to 128GB) — bought on Amazon 07/15/2026. Item $1,299 − $103.92 promo code − $115.25 gift card − $16.26 rewards + $56.31 tax = $1,119.88 all-in. Ships from Amazon to Honolulu, delivery Sun 07/19. Price watcher OFF — done.
✅ Purchased
GEEKOM A9 Max
Paid (all-in)
$1,119.88
Arrives
Sun 07/19
🔎 Live price watch Done · purchased
OutcomeBought the GEEKOM A9 Max on Amazon 07/15/2026 for $1,119.88 all-in ($1,299 list, less an 8% promo code, gift card + rewards).
Watcher🛑 Turned OFF 07/15 — no longer scanning prices; the machine is bought.
Deal alertClosed. No further price alerts.
Last checked07/15/2026 — final
🔌 Your setup: 1 monitor + share your laptop's keyboard & mouse
Status monitor✅ Yes — plug ONE monitor into the mini PC (HDMI/USB-C) showing your dashboards + what's running. All these boxes drive it easily.
One keyboard/mouse✅ Mouse Without Borders (free, by Microsoft) — glide your cursor from your laptop screens onto the mini PC's monitor and your keyboard follows automatically. No second set, no hardware, works over your network.
The "toggle"Either just move the mouse to that monitor (auto-switches), or set a hotkey to jump control between machines. Copy/paste and file drag work across both too.
Full control fallbackMicrosoft Remote Desktop takes over the whole mini PC in a window when you want it. Add Tailscale (free) to reach it from anywhere.
💰 Beelink SER9 Pro — 32GB Best value · $829
CPURyzen AI 9 365 (10C/20T, 5.0 GHz) + 73 TOPS NPU
RAM32GB LPDDR5x-8000 (soldered — cannot upgrade). Fastest RAM speed of any pick = best token/sec for local AI inference.
Storage1TB NVMe, second M.2 slot free (up to 8TB total)
Local AIRuns a 7–8B model comfortably alongside your full workload. 32GB caps you at that size.
Price$829 on sale (from $929) at bee-link.com · Win 11 Pro + 3-yr warranty included
Good used deal≤ $650 (eBay / Amazon Renewed)
Why pick thisCheapest way to offload your laptop NOW, and the fast LPDDR5x makes the most of 32GB for AI. If local AI outgrows it → sell + jump to Strix Halo. Flip cost ~$250.
🔧 GEEKOM A9 Max — 32→64GB Upgradeable · ~$1,170
CPURyzen AI 9 HX 370 (12C/24T) + 80 TOPS NPU
RAM32GB DDR5 SO-DIMM — upgradeable to 128GB. Buy + install 64GB kit (~$70) = 64GB for ~$1,170 total.
Local AIAt 64GB: runs 13B models comfortably. At 128GB ($200 more in RAM): runs 30B+ on-device.
Price~$1,099 (Amazon) / ~$1,175 (geekom.com w/ code GMSS16) + ~$70 for 64GB kit = ~$1,170
Why pick thisIf you want to scale the local AI without selling and rebuying. Upgrades to 128GB for ~$200 more in RAM — no new box needed.
🆕 Beelink SER10 Max — 64GB Just launched · $1,299+
CPURyzen AI 9 HX 470 (12C/24T, 5.2 GHz) + 86 TOPS NPU — newest gen
RAM32/64GB DDR5-5600, upgradeable (SO-DIMM). 10 Gbps LAN.
Price$1,299 (32GB) at bee-link.com · 64GB config still pre-order
Why pick / skipLatest gen, but $130+ more than GEEKOM A9 Max for similar specs. Wait for 64GB to ship + price to settle. Its launch IS pushing SER9 prices down.
🪙 Beelink SER8 — 32→64GB Budget entry · ~$550
The "test then upgrade" boxRyzen 7 8845HS, 32GB DDR5 upgradeable to 64GB. Runs your workload fine. Weaker for local AI (slower RAM, no NPU). ~$500–600 new.
Flip mathBuy $550 → sell ~$350 later → upgrade to Strix Halo $1,800 = net ~$2,000 (vs. $1,800 to buy Strix Halo direct = ~$200 flip tax).
💎 Strix Halo 128GB Endgame · ~$1,800
BoxesGMKtec EVO-X2 (~$1,800) or Framework Desktop (~$1,600–2,000)
Why it's the one16 Zen5 cores, 128GB unified LPDDR5X, ~256 GB/s bandwidth — runs genuinely capable local models (up to ~70B) on-device. HIPAA-safe AI by design.
When to buyWhen local AI on PHI is a validated core workflow, not a concept. Prices will be lower + more options in 6 months.
🧠 Your hybrid AI idea — validated
What you described is a legit, smart architecture. Here's how it works:
🏠 Local mini-AI (in the box)A small model via Ollama/LM Studio handles mundane, repetitive, and PHI-touching tasks — classify, extract, route, run automations. Data never leaves the PC → HIPAA-safe with no BAA.
☁️ Cloud Claude (heavy lifting)Full-context reasoning, creativity, judgment — on non-PHI only, until Anthropic BAA closes.
🔄 The bridgeCloud updates the local model's context/instructions when needed; local stays running 24/7 as the always-on worker. Exactly your idea.
What it needs⚠️ A useful local model = memory bandwidth + RAM → this is the whole reason the Strix Halo tier is worth it. On the SER9 you get a lighter local model; on Strix Halo you get a genuinely capable one.
📋 Specs you need — and why
| Part | Get | Why |
|---|---|---|
| CPU | 8+ core Ryzen (AI preferred) | Many Code sessions + automations running 24/7 |
| RAM | 32GB min · 64GB if upgradeable · 128GB for serious local AI | 32GB runs workload + 7B model; 64GB+ for bigger models |
| RAM type | Check soldered vs SO-DIMM | ⚠️ Soldered = locked forever. SO-DIMM = upgrade later for ~$70–200 |
| Storage | 1TB NVMe (2TB if storing PHI) | Room for local model files (~5–10GB each) |
| OS | Windows 11 Pro | ⚠️ Pro REQUIRED for Remote Desktop host + BitLocker. Home can't. |
| Security | BitLocker ON | Full-disk encryption = HIPAA baseline for any local PHI |
🔁 Pricing & used-buying guide as of 07/01/2026
| Model | New | Good used | RAM upgr? | Where |
|---|---|---|---|---|
| Beelink SER9 Pro 32GB | $829 (sale) | ≤ $650 | ❌ Soldered | bee-link.com · Amazon |
| GEEKOM A9 Max 32GB | ~$1,099 | ≤ $850 | ✅ to 128GB | geekompc.com (code GMSS16) · Amazon · Best Buy |
| Beelink SER10 Max 32GB | $1,299 | too new | ✅ to 64GB+ | bee-link.com (64GB pre-order) |
| Beelink SER8 32GB | ~$550 | ≤ $400 | ✅ to 64GB | eBay · Amazon Renewed |
| Strix Halo 128GB | ~$1,800 | buy new | ❌ Soldered | GMKtec / Framework direct |
FB Marketplace Oahu: search "mini PC", "desktop computer", "Beelink", "NUC" — small HI market but worth a check. Good-deal rules: ≤6 months old, original box, show idle temps. Avoid soldered-RAM units you can't verify the config on.
📡 Sales & releases in your 1–2 month window Watch
❌ Prime Day 2026Already ended 06/28 — missed it. GEEKOM A9 Max was $1,190 (from $1,399).
📅 Back-to-school salesJuly–August — Amazon/Best Buy/Beelink run PC deals. Best chance in your window.
📅 Prime Big Deal DaysTypically October — outside your 2-month window but worth knowing.
🆕 Beelink SER10 launch effectSER10 Max just launched → retailers clearing SER9 stock. The $829 sale price may drop further. I'm watching.
🆕 GEEKOM A9 Max (HX 470)New revision with HX 470 at $1,399+. May push HX 370 version lower.
I'll flagSER9 Pro under $800 · GEEKOM A9 Max under $1,000 · any strong Oahu used listing · Strix Halo price break
Full plan: KP Brain ▸ Resources ▸ Workstation Architecture. Flags I'll raise: price drops on either pick, strong used listings, CES refresh pricing, and a Win 11 Pro confirmation before you buy.
Ideas — how the team makes us better
🎤 Add your ideaAnyone can tap the 🎤 on a Process Maps task and speak an idea to make their work easier or automated. Sky replies, and the best ideas get built. This is everyone's board — see what your teammates are surfacing. 🔒 Non-PHI only.
Loading ideas…
Projects & tasks
Internal agency work, ranked and auto-blocked onto the calendar. 🔒 Non-PHI only — patient work lives in GoHighLevel.
Your calendar
Loading your day…
Custom fields
Activity
Activity
Comments open once this is created — hit Create and the thread starts here.
Collaborators
Status
Duration
Start date
Deadline
Saves automatically when you close
Subtasks
Each gets its own calendar block. Click one to open it — the parent completes when all are done.
Steps
Each step becomes its own task with its own calendar block, and can go to a different person. The project completes when all are done.
Start from a template — pre-fills the steps below (all still editable)
Subtasks
Add these once the task exists — hit Create and the ＋ Add button appears right here.
➕ Delegate a task
Brain-dump it — Sky turns it into clean task cards for the right person. 🔒 Non-PHI only.
Completed — permanent record, stays here even if removed from the calendar
Access admin
Add people and control exactly what each one sees. Changes save to Supabase.