Free to create an account. You confirm your email address with a code.
YouYou decide what matters
Needs YouDecisions come to you
ApprovalsApprove, reject, revise
RulesWhat your AI team may do
Your AI team in MasterMNGOrganized around your work
AI workersThe roles you decide
WorkProjects and tasks
KnowledgeWhat your AI team knows
Approved toolsPermissions and budgets
RoutinesWork that repeats
RecordEverything that happened
AI modelsThe intelligence each AI worker uses
ClaudeOther providers, once verified
You at the top, your AI team in the middle, the AI models underneath. How to connect your AI is on Providers.
Free account, open to anyoneOne company, your own Claude plan.
Structured handoffsWork passes between AI workers with its context.
Approvals checked at the serverDecisions bound to one exact action.
No model resaleYour provider bills you directly.
See it work
One objective. A team of AI employees. You decide when it matters.
You give your AI team an objective, the way you would brief a manager. It is split into work, done by the right AI employees, passed between them, and brought back to you when a decision needs a person. Then the work carries on, and every step is on the record.
An illustration of how work moves through MasterMNG. You assign the objective as a task; everything after that happens in MasterMNG's work, approvals and records.
You set an objective: "Research our competitors and prepare a launch plan."You
MasterMNG organizes it: your lead AI worker turns it into tasks for the teamLead
Strategy researches the marketStrategy
Marketing drafts the launch planMarketing
A reviewer checks the plan's claimsReview
Needs You: approval before anything is publishedYou
Work continues: your decision is recorded and the team carries onMasterMNG
The real software
This is MasterMNG: your AI team, working.
The Live Map shows your organization as it runs, from real activity: each AI worker inside its department with its real state — running, waiting, blocked, paused or failed — and what is waiting on you. A connection moves only while the activity behind it exists. Nothing is simulated to look busy.
What you are looking at: real MasterMNG screens from a demo company with demo data. Every product screenshot on this site is the real software; the diagrams are illustrations.
MasterMNG Demo company. Demo data; no AI worker is running in this capture.
For anyone with work to get done
Build the team you don't have yet.
Start with yourself. Create the AI workers your work needs, give each one a role, and grow the team as the work grows. The work defines the team — not your job title or the size of your company.
An architect might organize research, documentation, project coordination and operations.
A professional practice might organize research, documents, client work and administration.
A creator might organize research, content and partnerships.
A company might organize entire departments.
Illustrations of how a team can be organized: roles you create, not ready-made integrations for any profession.
You decide what needs doingYou
The team you need: AI workers with the roles you defineAI team
Your work: projects, tasks, knowledge and approved toolsWork
Results, on the record, with your decisionsResults
What MasterMNG is — and isn't
Not another AI chat window.
MasterMNG is the operating environment where AI workers have roles, work, knowledge, approved tools, permissions and accountability. The AI models provide the intelligence; MasterMNG provides everything around the team, so your work stays organized when the models change.
What it is
A place where your AI team works: roles, projects and tasks, knowledge, schedules and history.
Approved tools and permissions that you set.
Decisions that come back to you, and a record of what happened.
What it isn't
It does not replace human professional judgment.
It does not give AI unlimited authority.
It does not make every external integration available automatically.
AI workers act only within the permissions and execution paths MasterMNG actually supports.
The film · 1 min 44 s
The company operating system, in under two minutes.
From the Live Map and the Company Brain to work, knowledge, providers, a real recorded run and the decisions that stay with you. Every screen is the real product. There is no sound and no voice-over; the on-screen text, captions and transcript are in English only.
Product screens from our MasterMNG Demo company, with demo data. The execution and Decision Desk scenes are a recorded run from our internal demo company. Personal data is blurred.Read the transcript
0:00MasterMNG. The operating system for an AI-run company.
0:06The company at a glance: what is happening, what needs you, what finished.
0:12Live Map: every department and every agent, around the Company Brain.
0:20Company Brain: one graph of how the company operates.
0:27AI workforce: roles, managers, departments. A provider per agent.
0:33Work: epics to subtasks, on boards, sprints and releases.
0:39Knowledge: what the company learns, it keeps.
0:45Providers: your Claude plan on your own machine. Never a silent switch to API billing.
0:52Real execution: a Claude agent builds and hands off to a Codex agent. Recorded run.
1:01Verified, on the record: the reviewer runs the tests. 30 passed, 0 failed.
1:09Needs You: one queue for everything that needs a person.
1:15Decision Desk: nothing ships until you decide. The agent resumes once.
1:24Operations: routines, sprints and releases on one calendar.
1:30The whole company, on one map.
1:36Models change. Your company doesn't. MasterMNG, mastermng.com.
The text in this transcript appears on screen in the film, in English. On small screens a lighter 960-pixel version of the same film plays. The recorded run uses an internal demo configuration; the providers it shows are not a statement of which providers are available to customers.
Your AI team
Give every AI worker a role.
Create AI workers for the work you have — research, content, documents, engineering, operations, client work or whole departments. Give each one a role, responsibilities, work, knowledge, approved tools, permissions and, where it makes sense, a budget. They work on their own or together, each with a manager and a department, and you sit at the top. They support people; they do not replace professional judgment.
Departments, roles and reporting lines for every AI worker.
Each one shows its current task and the AI model it runs on.
When your AI team proposes a new hire, it waits for your approval.
MasterMNG Demo company, internal demo configuration. Demo data. Providers shown are not a statement of customer availability.
The work
Where the work gets done.
Your AI team works the way a good team does: on projects and tasks that are planned, assigned and tracked, with the result attached when they finish. For teams that want the detail: projects, epics, stories, tasks, bugs and subtasks, on boards, a backlog, sprints, releases and a roadmap. Every run stays on the task it served, failures included.
Your documents, decisions, procedures and know-how live in one controlled place that authorized AI workers can use, organized in spaces and pages with revisions, comments and templates. The Company Brain lets you see how your company is connected: AI employees, departments, projects, work, knowledge, rules, routines, decisions and activity.
Your AI employees research and analyse information, work on projects and tasks, create and review documents and files, use approved company knowledge, coordinate with each other, run recurring operations, hand work on to a colleague, and ask a person when a decision or an approval is needed. Everything they do is recorded.
A lead can split work across a team, combine the results and hand them to a person.
Your company's skills and rules apply to every piece of work they cover.
Every run is visible: running, queued, finished or failed.
Under the hood, AI employees work through the provider's own tools — Claude Code, on your own machine with the MasterMNG Runner. Other providers are listed on Providers once verified.
MasterMNG keeps work moving and brings you only the decisions that need a person: approve access, review a decision, resolve an exception, authorize a governed action. Everything waiting on people is in one queue, Needs You. When an AI employee needs a decision, it asks and stops. It carries on once, with your answer.
You decide what AI can do.
Permissions, approval policies and tool controls, set by you.
Approvals are checked at the server, and again when their effects are applied.
Every decision keeps who asked, from which task, and what you chose.
Capability policies, tool permissions where MasterMNG executes the call, server-side approvals and safe expiration. On your own machine, tool limits are applied by the provider's command-line tool, so enforcement there is partial. How it works, and its limits
MasterMNG · Decision Desk
The operations
Work that happens again doesn't need to be recreated every time.
Weekly reports, daily checks, month-end tasks: set them up once and they run on schedule, and you can see when one did not. Routines run recurring and one-off work on a schedule, with run history, pause and resume, and protection against running the same routine twice. The operations calendar puts routines, runs, task and epic dates, sprints, releases and decision deadlines on one calendar. Handoffs pass work between agents, or to a person, with the context and the run that produced it.
MasterMNG does not sell AI models. Your AI workers use your own Claude plan, on your own machine with the MasterMNG Runner. Each worker keeps its role, its manager, its work, its skills and its history in MasterMNG, so when the model behind it changes, the company does not. When one hands a task to another, the next one continues from a structured handoff — not from a transcript.
Product Engineer
Builds the calculator and writes the handoff.
Runs onAn AI model
Code Reviewer
Continues from the handoff and adds tests.
Runs onAn AI model
The company
Same task, same record, same decision for you.
Lives inMasterMNG
On the MasterMNG Runner, an agent set to use your plan is not moved to API-key billing.
We do not resell or mark up model usage.
A connection method is listed only once it is verified.
Write to contact@mastermng.com about what you need. This is an inquiry, not enrollment in an available beta programme. No additional access or delivery date is promised.
One company with AI workers, roles and departments, work, knowledge, the Company Brain and Live Map, routines, approvals, budgets and Needs You. You connect your own Claude plan. See the Product overview.
Which AI providers can I use?
Your Claude plan through Claude Code, on your own machine with the MasterMNG Runner. Other providers and connection methods appear on Providers only once they are verified for customers.
Do I pay MasterMNG for AI model usage?
No. Your provider bills you directly, and MasterMNG does not resell or mark up model usage. On the MasterMNG Runner, an agent set to use your plan is refused rather than switched to an API key if the machine is signed in with one.
Can AI workers act without me?
Only within the permissions you set. Governed actions wait on Needs You; approvals are checked at the server and bound to the exact action. On your own machine, tool limits are applied by the provider's CLI and enforcement is partial. Security explains the limits.
Is MasterMNG certified (SOC 2, ISO 27001)?
No. MasterMNG holds no security certification and does not claim compliance with any framework.
How can I discuss additional requirements?
Write to us from Discuss your requirements. An inquiry does not grant additional access or promise a delivery date.
Get started
Build the team you don't have yet.
Create your account, set up your company and give your AI workers their roles.
