Get useful business work without babysitting another chatbot.

Useful work first. A managed internal loop when it fits.

An AI workforce isn't one chatbot — it's a staffed org chart with bounded roles and approvals. You describe your business in plain English; the governed board turns that context into concrete internal work and routes each artifact to the right role. Here is the kind of first-round work three example businesses can generate. Configured schedules can prepare later internal rounds while outside authority remains with people.

Every action above is safety-checked and logged, and anything that spends money or can't be undone asks you first. Want it for your business? Brief the workforce below and watch it produce real work ->

Brief the workforce. Watch it produce real work, right here.

Not a video, not a mockup. Give the board a short business brief and the actual AI board composes the deliverables it would ship today - a staffed team, marketing copy, a founder to-do list, board decisions, a 90-day plan and more - tailored to what you typed. Nothing is saved; this runs the same engine your dashboard would.

A company name and description are enough. Add the customer and outcome for a sharper interview kit, roadmap, and revenue plan.

You can read every artifact above without an account. Create a free workspace and keep this brief -> The optional protected review above is a human-reviewed fit check, not enrollment, a quote, a purchase, or permission for bots to contact you.

What lands in your dashboard on day one.

Excerpts from the demo workforce. These are not mockups — the same copy ships from /api/dashboard?tenant=demo and into your own dashboard on signup, rewritten in your voice.

Six working pages — already live. Click any of them.

These are real development-preview surfaces, not screenshots. Every link below opens a working page. No sign-in is needed to look around.

A governed operating loop that keeps internal work moving.

One brief becomes role-owned plans, drafts, reviews, and receipts. Routine internal work can run on schedule; you remain responsible for decisions and every action that leaves the workspace. Here is the exact bounded loop.

What actually happens, every day

1. You brief the team — once.

   Briefing the AI team works exactly like onboarding a new executive hire: you answer 8 plain questions in a 5-minute wizard. No prompts to engineer, no config — just the same context you'd hand a new leader on day one:

   • What your company is, in one sentence
   • Your industry or category
   • Your mission — the problem you solve and why
   • Who your customers are
   • Your top 1-3 goals for the next 90 days
   • Your single biggest constraint right now
   • The voice your workforce should write in

   That brief becomes the company's standing context: every bot reads it before it acts, so you never repeat yourself. Change a priority and the whole team re-plans around it the next morning. See the exact questions →

2. The board and CEO turn it into a plan.

   Each morning the AI board meets, reviews yesterday's results, and the CEO bot breaks your goal into the day's work — a written standup with who owns what. You can read the full board minutes and the day's plan on your dashboard.

3. Work cascades down the org chart.

   EVPs route assignments to department directors, who route bounded tasks to staff roles. The 89 role definitions across 15 departments give each prepared task an owner, reporting line, and explicit permission boundary; the live loop activates work only as configured capacity and evidence allow.

4. Each task uses an allowed, measured route.

   The router can prefer plain code, a verified prior result, a local model, or another configured model when that route passes the task's evidence gate. Actual model, infrastructure, and review costs stay visible; the preview does not promise a cheapest route or call the workforce free.

5. Safety gates check every action before it lands.

   Before any result ships, it's checked against what that bot is permitted to do and logged so you can trace exactly what happened. Anything that spends money, signs something, or can't be undone is held and sent to you to approve first — nothing irreversible happens behind your back.

6. Finished work lands on your live dashboard.

   Brand guidance, marketing drafts, the 90-day plan, draft customer replies, and evidence-backed metrics show up on a dashboard that refreshes every 5 seconds. You see what was prepared, what's stuck, and who's working on what, then drop in tomorrow's brief. The governed internal loop can repeat without granting outside authority.

The 11 internal artifacts in the current demo manifest.

The public demo returns a labeled synthetic manifest so you can inspect the intended operating rhythm without mistaking it for customer work. In the private customer-zero workspace, scheduled internal work is written by named roles and logged to an audit trail. Outside delivery still requires approval.

• Board meeting minutes — chief of staff. The board convenes, works the agenda, and files one consolidated minutes record.
• Board directive — chief of staff. Sets the day's scope first, so every artifact below inherits a focus the board already ratified.
• Daily briefing — chief of staff. Yesterday's results, today's priority, and any open questions waiting on you.
• Board decisions — board leads. Three small, reversible calls logged with rationale — the AI exercising judgment, not just writing copy.
• Problem scout — strategist. Three concrete problems to probe today, each checked against your wedge.
• Customer interview kit — strategist. The actual script and wedge questions for the calls your TODO asks you to make.
• Marketing drip — growth marketer. One fresh, ready-to-post asset — an X thread, LinkedIn post, customer email, or launch teaser.
• Product build brief — web lead. One bounded proposed build, acceptance checks, and the source, verification, and release proof still required before anyone may call it shipped.
• Founder TODO — chief of staff. The five moves only you can make today, ranked and copy-pastable into any planner.
• Comments from the board — lane leads. One observation and one follow-up per lead on the work currently in flight.
• Debate lockout ledger — chief of staff. Every settled debate's verdict, locked so it reopens only on a new fact — never on repeat tension.

On Mondays a twelfth lands: the strategist's weekly plan. Nothing here is staged for the page — open the demo dashboard and you'll find today's eleven in the deliverables rail, each with its full body and the role that signed it.

What you get out of an AI workforce.

Home-Office is a workforce, not another app to babysit. Here is what that actually buys you — every line is something you can open and check on this live deploy.

• Coordinated capacity, not a labor-replacement promise. Stand up 89 governed roles across 15 functions to prepare routine internal work while people retain authority, judgment, and the benefit of the system.
• Internal work advances while you sleep. The board, CEO, and staff bots can run a scheduled preparation loop; your dashboard shows what was produced, what failed verification, and what still needs a person.
• Cost stays visible. A router prefers the lowest-cost allowed tool that passes the work's evidence gate; actual model and infrastructure costs remain measured instead of being called free.
• Nothing irreversible behind your back. Anything that spends money, signs, or can't be undone is gated and sent to you to approve first.
• Built for your business, not a template. Your playbook, brand voice, marketing copy, and 90-day plan are written for your company on day one.
• Auditable by design. Read the evidence, decisions, and protected-action receipts that prove what actually ran and what still needs work.

Built for operators who run lean.

If you're carrying more roles than you have people, Home-Office fills the gap with a working team — not another tab to manage. Here are five kinds of operator already in scope, and the specific work the AI workforce runs for each.

• Solo founder, pre-seed. You're CEO, marketer, and support all at once. The workforce writes your brand voice, launch copy, and 30/60/90-day plan, then prepares the next marketing draft and founder decision queue on schedule so you can stay on the product. Brief your team →
• Bootstrapped SaaS. No budget for a go-to-market hire. Sales prepares approved lead research and outreach drafts, Customer Success organizes connected support signals, and Data summarizes only verified metrics. See the dashboard →
• Agency or studio. Every client wants senior output on a junior budget. Spin up a full 89-role org per brief, keep the brand voice and deliverables auditable in the ledger, and self-host it under your own keys. Meet the roster →
• Ecommerce or DTC brand. Marketing and support eat the week. The workforce drafts product copy, launch material, and suggested customer replies in your voice; a person approves customer contact, spend, and refunds. See the daily output →
• Side project, nights and weekends. Prepared internal work can keep moving while you're at your day job. The scheduled loop organizes drafts, evidence, and decisions; anything outside the workspace waits for you. Read the loop →

89 AI workers, each with their own job.

From CEO to clerk across 15 departments — every AI worker has their own role, knows what they're allowed to do, and remembers their past work. Here are eight of them and the features they actually run for you.

Chip · CEO

Daily standup + strategy. Reads your goals, splits the day's work across the team, and posts a morning brief to your dashboard.

Penny · CFO

Budget + invoicing. Tracks every dollar your bots spend, drafts invoices, and flags any spend you should approve before it goes out.

Merge · CTO

Code review + verification. Prepares code changes, runs the test suite, and holds every merge or production change for explicit approval.

Buzz · CMO

Brand voice + landing pages. Writes your marketing copy, your homepage hero, and your launch announcements in your own brand voice.

Pip · CRO

Pipeline preparation. Organizes approved lead research and drafts outreach for review; it never contacts a prospect or books a call on its own.

Cami · Customer Success

Support preparation. Organizes connected support signals and drafts replies in your tone; a person approves every customer message.

Lex · General Counsel

Contracts + privacy review. Reads every contract, redlines the risky clauses, and keeps your privacy policy current.

Dot · Data Analyst

Evidence-backed metrics. Summarizes connected, verified revenue, signup, and churn evidence; missing data stays visibly missing.

15 departments, each with a bounded role definition.

The 89 role records are organized into 15 departments so work has an owner and an approval boundary. This is an operating model under customer-zero testing, not proof that every department continuously runs an outside business. Roles prepare internal work; people approve what leaves.

• Executive. Organizes priorities and prepares a board brief for review.
• Product. Proposes roadmap changes and drafts specifications against your goals.
• Engineering. Prepares code changes, reviews them, and runs tests; merging remains gated.
• Design. Prepares brand, layout, and visual drafts for human approval.
• Marketing. Drafts copy, landing pages, and launch posts without publishing them.
• Sales. Prepares lead research and outreach drafts without contacting prospects.
• Customer Success. Organizes account signals and drafts onboarding or retention plans.
• Support. Drafts replies and escalation context without sending customer messages.
• Finance. Summarizes verified financial evidence and drafts invoices; it cannot move money.
• Legal. Flags contract and policy issues for qualified human review; it does not give legal approval.
• Operations. Triages internal queues and prepares chief-of-staff receipts.
• Data. Produces aggregate rollups from available evidence without inventing missing metrics.
• Security. Checks allowed actions, retains evidence, and requires human review before external submission.
• HR. Defines each role, what it reports to, and what it's permitted to touch — the permissions that keep 89 bots from stepping on each other.
• IT. Routes bounded tasks and monitors the dashboard and approved integrations.

Every department has role records in the live customer-zero roster — open the 89-role gallery and filter by any one to inspect the role definitions.

How much the AI does on its own — we level up carefully.

Assigned and scheduled internal work can run inside the workspace. Strategy, external communication, publishing, contracts, spending, and anything irreversible remain gated, one evidence-backed step at a time.

Current stack. Verified boundaries.

Home-Office currently connects Cloudflare Pages to a private Supabase workspace and keeps local-model routing available for bounded work. Here is what's wired today — and what's honestly still on the roadmap.

• Governed model routing. The customer-zero router sends bounded work only through configured model credentials and records which route ran; preview access does not promise every provider integration.
• Local model option. A configured local SLM can avoid per-call API charges for suitable work; electricity, hardware, network, and review time still count as delivery cost.
• Supabase for private workspace state. Founders, tenants, and AI role records use canonical Postgres tables rather than a browser-only session. Export scope and customer responsibilities remain part of the review.
• Verified deployment. The current customer-zero build ships as static pages plus functions on Cloudflare, with Supabase-backed private workspaces.
• Scriptable operating loop. An auditable action ledger and plain configuration let TIMxAI connect authorized bots to reviewed tools.
• App connectors on the roadmap. Native Slack, Stripe, email, and calendar hooks are in active development — we list them as roadmap, not a live promise, and won't fake a logo we haven't shipped.

Evidence first. Claims second.

Home-Office is in customer-zero development. No source-code license, commercial access, support promise, or hosted capacity is granted by this page. The product advances only when its operating evidence passes the mission, roadmap, build, release, support, and outcome gates.

• Auditable by design. Every action a bot takes is logged with the role, tool, and budget — read it like a ledger.
• Local-first router. Cheap models handle the routine; premium models are only paged in when a task actually needs them.
• Protected decisions. Money, contracts, external communication, and irreversible actions remain human-gated during customer-zero proof.

The case study is this company.

No borrowed logos. No invented five-star reviews. Home-Office is early, and instead of faking proof we let the AI workforce build and run TIMxAI — the company behind this product — in public. Every claim below links to something you can open and check yourself.

Named customer case studies will appear here as Home-Office opens to outside founders. Until then the proof is public and auditable: open the demo dashboard, read the board minutes, and watch the workforce ship in real time.

We run the operating loop. You keep the authority.

The intended paid product is not access to a chatbot or a dashboard sticker. It is managed product stewardship: TIMxAI connects the evidence, keeps routine internal work moving end to end, verifies the output, and brings you a short decision queue for the actions only you can authorize. The free preview exists so this value can be judged before any price exists.

• Intake and priorities. Turn source-linked product, support, incident, and outcome evidence into one ordered backlog instead of asking you to repeatedly prompt an AI.
• Prepared work and verification. Produce bounded code, content, operating, and release artifacts; test them; retain receipts; and hold anything that fails.
• Monitoring and follow-through. Watch connected product surfaces, triage support signals, measure outcomes, and feed the evidence into the next recommendation.
• One human decision queue. Customer contact, publishing, spending, contracts, production changes, security submissions, and unclear authority remain explicit human decisions.
• Weekly operating truth. Report work completed, failures, owner minutes, delivery cost, product outcome, and verified revenue without presenting activity as success.
• A human-centered North Star. TIMxAI is building toward equal weekly pay for a human-approved cohort of people impacted by AI, after verified obligations. Today distributable income and the approved cohort are both zero; this is the direction, not a funded-result claim.

Current boundary: this managed service is still a review draft, not an offer or SLA. No price, checkout, outside managed-service enrollment, autonomous external communication, or revenue claim is active. Read the managed operating scope →

Free preview. No price. No checkout.

Create a private development workspace free and judge the work it produces. It is not for sale as a paid managed service; TIMxAI is proving the operating loop before asking anyone to buy it.

There is no card, founding price, subscription, sales call, implied license, or future obligation. Useful operating lessons remain free.

Questions founders actually ask.

No spin. Each answer points at something on this live deploy you can open and check for yourself.