A goal with an owner
Each workspace is built around one outcome a person can check, such as "client update approved by Friday." Unrelated work lives in its own workspace so instructions and sources stay clear.
Keep the documents, AI agents, instructions and results for a recurring job together. Start with a Manor template, connect supported tools and configure review rules. For the next report or client update, add current inputs and reuse your setup.
An AI workspace brings source material, agents, instructions and outputs into one place. In Manor, you can save the setup for a business task, choose the sources for each run and inspect the result and activity history. Your permissions and configured review rules determine which actions can run.
Each workspace is built around one outcome a person can check, such as "client update approved by Friday." Unrelated work lives in its own workspace so instructions and sources stay clear.
Approved files, threads, notes, schedules, and earlier outputs stay attached to the job. Agents work from them instead of whatever someone remembers to paste into a chat.
Drafts show which source each claim came from, gaps are flagged instead of guessed, and an activity log records what ran and, when applicable, who approved it.
Choose a specific job below, or compare the agents, knowledge, tools and review features available in a workspace.
Bring a Gmail connection, service notes and reply rules. Prepare drafts and follow-up tasks that you can review against those sources.
Set up an email agentBring this week's notes and a reporting format. Prepare progress, blockers and next steps, then reuse the instructions with new inputs.
See a weekly update workflowBring a brief and the material the presentation should cover. Generate an editable PowerPoint file and check it before sharing.
See presentation generationHere is how a small agency could run its weekly client update in one workspace. The scenario is illustrative, not a customer case study. Our client delivery workflow guide walks through a similar studio project in more detail, including acceptance criteria and a pilot checklist.
The signed SOW or brief, the client email thread, call notes, the project tracker, and last week's approved update. One person keeps these current.
Agents compare this week's tracker and threads against last week's update, draft the new update with a citation for each statement, and flag claims the sources do not support, such as a launch date nobody confirmed.
In this example, the account owner is configured as a reviewer. They read the draft and the flags, then approve, edit, or leave comments. Nothing goes to the client until it passes this approval step.
Next week's scheduled run reuses the saved reporting instructions. The team selects or adds the latest sources and can include the previous draft and owner's feedback as needed, so a note like "lead with blockers" or "don't mention the staging link" does not need to be repeated.
The same pattern fits other recurring jobs. For an online shop, an AI workspace for ecommerce operations turns dated stock exports into a weekly exception brief the owner checks before any purchasing decision.
These tools often work side by side. The difference is what each one keeps and who decides. For a longer comparison, read AI workspace vs automation builder.
| Question | AI chat assistant | Automation builder | AI business workspace |
|---|---|---|---|
| Primary job | Answer a question or draft text now. | Run predefined steps when a trigger fires. | Keep a recurring business job moving toward its goal. |
| What carries over | The current conversation and files you paste. | Only the data passed into each step. | Goal, approved sources, past outputs, open follow-ups, and reviewer feedback. |
| How decisions are made | Generate the most useful reply. | Follow fixed branches and conditions. | Agent judgment inside the permissions and approval rules you set. |
| Best fit | One-off research, writing, and questions. | Stable system-to-system processes. | Work that combines reading context, drafting, routing, and human review. |
Begin with one job a person can review in minutes. If you are comparing products, our guide to the best AI business workspace for small business lays out the options by use case.
Choose work that repeats, such as a client update, inbox triage, a follow-up queue, or a weekly report. Write down what "done" means and who approves it.
Add the approved documents, threads, and tools for that job and nothing else. Missing access should show up as a visible blocker, not a guess.
Let agents prepare summaries, drafts, and internal tasks. Keep sends, money, commitments, and account changes with a person.
Check that citations point to the right source, gaps are flagged, and review takes less effort than doing the work by hand. If not, simplify the scope first.
Plans and usage are shown in the Manor app before you start a workspace. The FAQ covers current pricing questions.
You do not have to begin with an empty workspace. The AI workflow guides library documents common jobs with their inputs, outputs, and start conditions. Installable workspace Blueprints are listed in the public Manor Workspace Marketplace; an installed Blueprint becomes an editable copy in your own account, and you connect your own data and tools. A creator's private workspace, credentials, and customer data are not transferred. If you have a method that works, you can publish it as a Blueprint.
The hosted app at app.manorai.xyz, where you create workspaces, install Blueprints, and connect the tools available to your account.
Manor's source-available self-hosted edition lets teams inspect the architecture and operate their own workspace. It uses the Sustainable Use License 1.0, not an OSI-approved open-source license.
An AI business workspace is a shared place where one recurring piece of business work lives: its goal, approved sources, agents, drafts, and decisions. It retains the setup and prior results while a team can replace current inputs and select sources for each run.
Yes. AI workspace for business, business AI workspace, and AI business workspace describe the same idea: AI agents working on ongoing business jobs from shared context, with people reviewing the results.
A chat assistant answers the question in front of it. A workspace keeps the goal, sources, schedule, open follow-ups, and past decisions for a recurring job, so the next run continues the work and a reviewer can see where each statement came from.
An automation builder runs fixed trigger-and-action steps you define in advance. A workspace lets agents read documents, draft, and flag gaps inside the permissions you set. Many teams keep deterministic automations for system-to-system syncs and use a workspace for work that needs reading, drafting, and review.
No. Keep your CRM, accounting, inventory, or project system as the system of record. The workspace reads from approved sources and coordinates the review and follow-up work around them.
Actions follow the configured permissions and review rules. For sensitive sends, payments, refunds, pricing, contracts, and account changes, we recommend configuring review before relying on the workflow.
Yes. Browse the Manor Marketplace for installable workspace Blueprints. An installed Blueprint becomes an editable copy in your own account, and you connect your own data and tools. Workflow Guides explain the sources, setup and review steps for configuring your own workflows.
Manor Cloud lists Free, Personal at $20 per month and Pro at $49 per month, with custom Enterprise terms. Paid templates are separate one-time purchases. Check included credits and any additional model or tool charges in the app. Self-hosting has its own infrastructure and model costs.
These guides are illustrative setups. Use them to identify the inputs and review steps for your work, then confirm the supported tools and template requirements in your account.
Find a workspace Blueprint in the Marketplace, check its setup requirements, then install your own editable copy.