04 · Company Install
A brain for every employee. One that ties them together.
Custom scope · 2–6 month engagement. Minimum 3 Personal Brains required.
SMBs, departments inside larger orgs, mid-market companies installing AI org-wide. Leaders who care about the line between individual contribution and institutional memory.
Each person gets a brain that’s theirs to keep. The company gets a shared brain that holds the institutional knowledge. Both work together. Both honor the line between what’s personal and what’s shared.
The structural problem with company-wide AI.
Most companies installing AI right now are making the same architectural mistake. They’re either centralizing everything (one shared workspace, one chatbot, all knowledge funneled into the corporate vault) or decentralizing everything (let everyone have their own ChatGPT subscription and figure it out). Both approaches break in predictable ways within twelve months.
Centralize-everythingfails the employees. The knowledge people have built up over years of doing the job lives in their heads, their email, their personal Drive, their habits — and most of it can’t be moved into a corporate AI without losing the personal context that made it useful. People resist contributing because what they contribute becomes the company’s property the moment they upload it, and they correctly intuit that this affects their leverage. The shared brain ends up half-empty.
Decentralize-everythingfails the company. Each person builds their own little ecosystem of prompts and projects. The institutional knowledge that should compound across the team doesn’t — because there’s no shared layer. When someone leaves, their AI work product leaves with them. When two people in the same company need to coordinate on a shared problem, they’re each working with a different mental model and a different AI. The aggregate effort is wasted.
The Company Install resolves this by being explicitly two-tier. Every employee gets their own Personal Brain — owned by them, portable when they leave, fed by their own work and judgment. The company gets a shared Company Brain — owned by the company, fed by what the company has explicitly decided should be institutional knowledge. The two query each other through a governance layer that respects the line.
What the architecture actually looks like.
Each Personal Brain is the same architecture as the solo Install: branded subdomain, gated by employee identity, MCP-queryable from any modern AI tool, with role-tuned skills installed inside the workflows the employee already runs. The corpus is built from the employee’s work, with their consent, structured around the role they actually hold.
The Company Brain is a parallel architecture for institutional content — policies, playbooks, frameworks, case studies, training material. It lives on a separate subdomain, gated by company-wide SSO (M365, Google Workspace, or Okta), queryable by the same tools that reach individual Personal Brains.
The connective tissue is the governance layer. Most installs use a three-tier model. Personal → Personal queries are off by default; cross-brain rights only open when both employees opt in. Personal → Company queries are on by default, scoped by permissions. Company → Personal queries are off entirely — the company never automatically reaches into individual brains.
Two-tier architecture · personal owned, company owned
This is the doctrine. The Personal Brain holds what walks out the door when someone leaves. The Company Brain holds what stays. The governance layer makes both legible.
“The company brain holds what stays. The personal brain holds what walks out the door. The architecture honors both — because pretending one of them doesn’t exist is how AI rollouts fail.”
The architecture is the half that matters
Scope it on a call before you scope it on a contract.
Diagnostic Call (30 min, free) maps your team to the right tier and outlines the governance contract. No commitment.
Four tiers, sized by team + data scope
Pick the tier that matches today.
Small · 1–15
$15K + $1.5K/mo. Company Brain + 3 role-skill types + governance + cross-brain rights. First organized AI footprint.
Mid · 15–50
$25–45K + $2.5–3.5K/mo. Adds multi-department governance, onboarding curriculum, audit logs, M365/Google/Slack.
Large · 50–250
$50–100K + $5–7.5K/mo. Multi-tenant platform, cross-department rights, structured 6-month rollout program.
Enterprise · 250+
Custom. Full institutional embed. 12+ month engagement. AI rollout integrated with operating-model change.
What you walk away with
In the bundle.
Two-tier brain system
Personal Brains for enrolled employees (owned by them). Company Brain holding institutional knowledge (owned by the company).
Governance contract
Written, agreed-upon, technically enforced. Who can query what, who keeps what when they leave. Half the value of the install.
Role-tuned skill library
Shared skills installed across brains, tuned to the function. Sales teams get sales-shaped, engineering gets engineering-shaped. Compounds across the team.
Stack integration
M365, Google Workspace, Slack — your daily tools query and feed the brains. Mid tier and above.
60-day support
Structured stabilization window after launch. Most installs hit a "the architecture didn't imagine this edge case" moment in week 3–4; this catches them.
Docs employees read
Not vendor-speak. Written in the same voice as the rest of the site — diagnosis, not promise. Explains what each piece does, why it\'s there, how to use it.
Why "minimum 3 Personal Brains" is the rule.
You can’t have a Company Brain without people who’ll actually use it through their own. Every install that violated this rule — companies who tried to deploy just the shared brain — produced the same outcome. The shared brain becomes a write-only archive. Employees keep using their personal ChatGPT and Claude because that’s where their work flow is. The investment gets shelved within a year.
The fix is to deploy from the user out. Each employee’s Personal Brain gives them an immediate reason to engage. The Company Brain then becomes useful because they’re already querying through tools that can reach it. The shared brain compounds because the personal brains are pulling from it constantly.
Three is the floor because below that the cross-brain governance design isn’t worth the architectural complexity. At three, the shared layer starts paying for itself. At ten, it’s indispensable.
Who this is for
Named by role.
- The founder of an SMBYou've got 10–40 employees. You want AI rolled out properly the first time. You don't want to learn the hard way which pieces of architecture matter.
- The COO or Chief of StaffYou own operational rollouts. The board is asking about AI strategy. You need a real plan, not a vendor list.
- The department head inside a larger orgYou can run a real install at the department level while corporate IT is still picking a vendor. The architecture is portable to whatever the company eventually settles on.
- The Head of People / HR leaderYou care about the employee-ownership side. The portable Personal Brain story is what makes AI rollout an upgrade for employees, not a surveillance threat.
What’s included
In the price.
- Discovery + scoping → architecture design → per-employee Personal Brain installs + Company Brain build
- Personal Brains owned by each employee (portable when they leave the company)
- Company Brain owned by the company (stays through employee turnover)
- Role-tuned skill library for the function the team performs
- Governance overlay — who can query what, cross-brain rights, audit trail
- SSO integration via M365, Google Workspace, or Okta (mid tier and above)
- Optional: training sessions, M365 / Google / Slack deeper integration
- 60-day post-launch support window
- Written governance contract employees can actually read
- Minimum 3 Personal Brains required (Company Brain alone is not a valid configuration)
Common questions
Asked, answered.
What happens when an employee leaves?
They keep their Personal Brain. They walk out the door with the corpus, the schema, and the deployment — same as if they bought the install themselves. The Company Brain stays with the company, untouched by the departure. The governance layer revokes their access to query the Company Brain the moment their employment ends.
This is the design, not an accident. The portability is what makes employees willing to invest their judgment into building their Personal Brain in the first place. The institutional separation is what keeps the company’s knowledge intact when people turn over. Both honor the line.
Doesn’t this create a security risk — employees walking out with company data?
No, because the Personal Brain’s corpus is built from the employee’s own work, with their consent — not from automatic ingestion of company files. The Company Brain holds the institutional knowledge that the company has explicitly decided is institutional. Employees can’t download the Company Brain into their Personal Brain — the governance layer prevents that. What an employee can take is what they’ve personally produced, which is what they could already take anyway (in their head, in their drafts, in their personal files). The Install just makes it structured and portable.
How do you handle regulated industries?
The architecture supports HIPAA, SOC 2, and similar compliance frameworks because it’s built on Cloudflare’s enterprise infrastructure with explicit access gating. For regulated installs, we add a compliance scoping phase during discovery — usually a 1–2 week extension to the timeline — where we map your specific requirements and design the audit trail accordingly. For installs that touch PHI or PII at scale, we may bring in a compliance specialist to review the architecture before deployment. Flag your industry during the Diagnostic Call.
Can you integrate with M365 / Google Workspace / Slack?
Yes, at the Mid tier and above ($25K+ engagements). The integration covers SSO (so employees use their existing company login), document ingestion (so files in shared Drives can feed the Company Brain), and conversational surface (so employees can query the brain inside Slack or Teams). Small tier installs can add integration as a follow-on engagement once the architecture is in place.
What if my company already bought Microsoft Copilot or Google Gemini Enterprise?
The Company Install layers on top of whatever AI tools you’ve already deployed. The Personal and Company Brains expose MCP endpoints — both Copilot (via custom plugins) and Gemini (via extensions) can query them. The Install doesn’t replace your vendor tools; it gives them better context to work with. Most companies that buy Copilot find that the value lands once their Personal and Company Brains are wired in — the bare Copilot answer is generic; the Copilot answer that’s read your company’s playbooks is specific.
How much custom development is involved?
The infrastructure is productized — Workers, D1, Vectorize, Workers AI, Cloudflare Access — and built off a shared template that gets configured per install. The custom work is in (a) the corpus extraction for each Personal Brain, (b) the architecture design that matches your data scope and governance needs, and (c) the role-tuned skills for the function. We aren’t writing a custom platform from scratch for each company — we’re configuring a platform that already exists, against the realities of your team.
What's the timeline?
Small tier: 2 months from kickoff to launch. Mid tier: 3–4 months. Large tier: 4–6 months. Enterprise: 6–12+ months depending on scope. The variance is mostly in the discovery and architecture phases — the actual brain installs run roughly 4–6 weeks each, but they can be parallelized once the architecture is set.
What about ongoing costs?
The monthly retainer covers ongoing care: corpus updates, governance changes as the company evolves, new skill installs, response to break/fix. The infrastructure pass-through cost (Cloudflare + Twilio + LLM tokens, if applicable) typically runs $50–$500/month depending on team size and usage. The math is dominated by the value of the time saved — for a team of 20, a brain that saves each employee 30 minutes a day pays for the entire engagement within a quarter at any reasonable hourly cost.
Adjacent paths
Where this connects.
Start with the Diagnostic Call.
The Company Install starts with a 30-minute Diagnostic Call. Free. Scoped to figuring out which tier matches your team, what the governance line should be, and whether the install is the right move right now or whether something smaller (an Audit, a Workflow Sprint) is the better first step.
As featured in
