Updated Jul 2026 · Care Plan + hardware kits

Private AI help, set up for real people.

Done-for-you local-first AI for families and older adults — free fit check, Home Setup from $497, optional Care Plan $79/mo, and hardware that matches how offline (and how open-source) you want to be.

Proof of concept: a 68-year-old mother who always wanted to write her memoir started immediately with her AI writing assistant — then called to say thank you for the motivation, focus, and beautifully edited chapters.

Fit check first Care Plan $79/mo Hardware offline ladder

Built for families who want AI without becoming tech support.

The first customer is often an adult child, caregiver, teacher, or community helper who wants safe AI support for someone who does not want a complicated setup.

For older adults

A gentle helper for drafting messages, organizing thoughts, explaining confusing text, staying on task, and making creative projects like memoir writing feel possible.

For adult children

A safer path than handing a parent a random chatbot account and hoping they understand privacy settings, scams, and limitations.

For community helpers

A practical way to bring local-first AI support into classrooms, churches, nonprofits, coaching, and small organizations.

What we set up.

Phase 0 is intentionally human-led. We learn what breaks, keep the safety steps careful, and only automate what proves repeatable.

Readiness check

We confirm the Mac or Windows PC, internet connection, permissions, and customer goals before changing anything.

Local AI brain

We install or verify Ollama and a practical local model so the assistant can run on the customer’s computer.

Friendly assistant profile

We configure a simple assistant personality with plain-English safety guidance and realistic boundaries.

Optional phone chat

If requested, we connect a familiar chat app and explain clearly which parts are local and which are not.

Handoff and support

The customer sends a test message, receives a handoff guide, and learns how to turn remote support off.

Honest privacy beats hype.

Offline Helper should never claim “100% offline” when phone messaging, cloud backup, email, or other internet services are used. The promise is local-first, careful setup, and plain-English boundaries.

Safety boundaries

  • No medical, legal, financial, crisis, or emergency decision-making.
  • No public exposure of the local AI model.
  • No passwords, bot tokens, or recovery keys pasted into documents or chat logs.
  • Remote support requires explicit consent and can be turned off.
New option · Odysseus self-hosted workspace

A private ChatGPT-style workspace, installed carefully.

Odysseus is a free, open-source, self-hosted AI workspace for chat, agents, research, documents, notes, email, calendar, and local model workflows. Offline Helper can now evaluate whether it fits a household or small organization, install it safely, connect a local model such as Ollama, and explain the privacy boundaries in plain English.

Guardian setup defaults

  • Auth stays on; no public exposure of the workspace or raw model ports.
  • Local-first where practical, with clear labels for any cloud/API features.
  • Family-safe handoff: passwords, backups, updates, and support notes documented.
  • Plain-English walkthrough for memoirs, letters, research, planning, and learning.

Fit check first. Setup once. Care Plan monthly.

Offline Helper is a service business, not a chatbot subscription. You get a working private AI on hardware you control, then optional monthly care so it keeps working — plus curated hardware if you want deeper offline / open-source hosting.

Step 1 · free

Private AI Fit Check

Free 15–20 min

We confirm goals, device readiness, privacy comfort, and which offline / open-source tier is realistic before anyone pays.

  • Who the helper is for
  • Mac or Windows readiness + memory
  • Offline depth vs cloud boundaries
  • Honest model-size recommendation

No payment before fit check. If it is not a good fit, we say so kindly.

Book a free fit check
L1 · loved ones

Family Care Setup

$997 setup

For adult children, spouses, and caregivers who want the setup to feel safe, dignified, and easy to support.

  • Everything in Home Setup
  • Caregiver orientation
  • Consent + support-boundary sheet
  • 30-minute follow-up session

Payment only after fit check. Consent path must be clear before checkout.

Set up for a loved one
L2 · keep it working

Care Plan

$79 / month

The product that keeps private AI useful after install day — updates, health checks, and one new skill each month.

  • Model / software update check
  • Backup & archive health (when configured)
  • Skill of the month
  • One short support call
  • Privacy / scam boundary refresh

Cancel anytime. Usually attached at setup or day 14. Offered after value is proven.

Ask about Care Plan
Also available: Micro-Org Setup ($1,500) for 1–3 seats / partner sites · Paid Rescue Session ($99, 90 min) when something is stuck today · Request rescue · Hardware kits ↓

How offline — and how open-source — do you want to be?

Hardware and network kits set the ceiling: soft offline on a laptop, mostly-offline on a home node, or deep offline with a GPU lab that can host larger open models. Fit check first; we never upsell hardware you do not need.

H1 · soft offline

Edge Privacy Kit — $349

Network and device hygiene on the computer you already own. Improves privacy posture without buying a new box.

Offline depth: soft offline (local model when RAM allows)
Open models: ~3B–8B on existing hardware
Includes: guest Wi‑Fi / IoT isolation guidance, DNS hygiene, no open ports, plain-English boundary map
Discuss on fit check
H2 · mostly offline

Home AI Node — $1,499

Curated mini PC plus home LAN setup so the family AI lives on a box you control — not a random cloud tab.

Offline depth: mostly offline on the home LAN
Open models: ~7B–14B daily drivers
Includes: sized mini PC, imaging, local chat UI, network isolation, handoff training
See if Home Node fits
H3 · deep offline

Open Model Lab — $3,997

GPU workstation class for power users and small orgs who want to self-host larger open-weight models.

Offline depth: deep offline / air-gap-capable option
Open models: ~32B–70B quantized (VRAM dependent)
Includes: curated GPU stack, cooling/power notes, local UI only by default, staff handoff
Plan a lab build

Honest limit: “100% offline forever” is rare if email, phone chat, cloud backup, or voice services stay in the mix. We size kits to your dream outcome and electricity/noise budget — not hype. Shipping, tax, and exact SKUs are confirmed after fit check.

Research loop before we overbuild.

Every Facebook reply, fit check, paid setup, abandoned checkout, and support question becomes a redacted UX/UI signal. We improve copy, pricing, accessibility, and automation only after patterns repeat.

Capture consented signals

Use a short consent note and survey. Do not collect passwords, tokens, or private documents.

Validate privacy

The research script checks for missing consent and obvious secret-like text before reports are generated.

Pick one UX change

Weekly reports identify top friction, price reaction, accessibility needs, and the next experiment.

Free community

Join the Offline Helper community before you book.

Want examples, safe setup notes, and plain-English local-first AI guidance first? The free Skool community is a soft place to learn, ask general questions, and see whether Offline Helper feels right before a fit check.

Privacy note: community posts should stay general. Please do not share passwords, private documents, recovery keys, medical/legal/financial details, or crisis information there.

Start with a careful fit check.

Offline Helper is a human-led setup service for families, community helpers, and small organizations that want practical AI help without handing over passwords or private documents.

We begin with a free fit check so we can confirm your device, goals, privacy needs, and whether a local-first setup is the right path before you pay for anything.

1. Fit check first

Confirm goals, Mac or Windows PC readiness, consent, and local-versus-cloud understanding before asking for payment.

2. Review the setup plan

If it is a good fit, you receive a plain-English plan, price, schedule, and secure payment option before any setup work begins.

3. Set up and hand off

We configure the tools with you, explain what changed, and leave you with simple next steps you can actually use.

Start with a free fit check.

Best fit: a Mac or Windows PC and a family who wants private AI without becoming technical. Fit check first — then Home Setup, Care Plan, or a hardware kit only if it matches your offline goals.