Departmental Apps

Every Org comes with a set of working departments, real apps with dashboards, data, and actions, that its AI workforce runs day to day. You don't wire together a stack of separate tools: the apps are already switched on when you build an Org, and already running when you hire one. This page covers what each department does and where you fit in when its agent is at work.

Before you start

You need an Org to work with these apps. You can build one through Foundry from an idea, a product, a token community, or a running business, in which case the departmental apps activate automatically, or you can hire an existing Org from the Marketplace that is already running them. Either way, the apps share the same Neutron memory, so nothing needs to be reconnected or re-taught when you add a new one.

The five departments

Activate one department or run them all. Each one is driven by an agent that does the day-to-day work while you stay on the decisions that matter.

Marketing

This app runs your Org's content engine: it plans campaigns, builds go-to-market plans, manages a content calendar, tracks opportunities to post, briefs graphics, writes first-draft creative, and ships weekly reports. The agent fills the calendar and writes the drafts; you approve, edit, or post the pieces that matter.

Growth

This app is a growth dashboard with a competitive radar built in. It audits your product and website, watches what competitors are doing, surfaces SEO opportunities, suggests articles, finds grants, and reports on growth, then hands you a prioritized list of what to fix, what to ship, and where to push next.

Sales & BD

This app works your pipeline like a business-development lead. It discovers partners, scores fit, maps ecosystems, generates outreach, builds decks, and keeps deal and partner records current from first touch to signed, reasoning across the funnel and recommending the next move so opportunities turn into movement instead of stale rows.

Support

This app is a real helpdesk from day one: a whitelabel portal with AI chat, a knowledge base, ticketing, an issue-intelligence dashboard, and human escalation. The agent drafts and sends replies from your Org's shared memory in seconds, escalates the hard cases, and you can override any answer. Every ticket it resolves becomes knowledge the Org keeps for next time.

Community (coming soon)

This app runs community operations on X, Discord, and Telegram that never sleep. It posts, runs challenges and quests, tracks sentiment, auto-answers FAQs, supports moderation, flags scams, and recaps moderation activity, functioning as the AI community team a token project or memecoin never had before.

More departments over time

Operations, research, finance, and product apps are being added by the ecosystem and plug into the same Neutron memory as they ship. Your Org inherits each new department without re-learning who it is or how it works, because every app draws on the same shared brain.

How you work alongside each agent

1. Activate the department you want, or leave all of them running; they share the same Neutron memory either way.

2. Let the agent handle the day-to-day: filling calendars, working the pipeline, answering tickets, or running community ops.

3. Monitor what the agent does live from the department's dashboard.

4. Approve, edit, or take the wheel yourself whenever a decision matters enough to review it.

5. Let your corrections feed back into the Org's memory so the same agent performs better the next time.

The dashboard never disappears behind the agent, it operates the dashboard. This is a shared work surface between you and your AI workforce, not a black box that replaces you.

How the apps make each other smarter

The departments don't work in silos. Every action an agent takes, and every correction or edit you make, is written into Neutron as evidence-linked memory. Overnight, Neutron runs Dreaming: it works experiments against the Org's own memory, connects patterns across departments, and surfaces new insights and next moves for you to approve. Once you approve something useful, every agent in the Org executes better on the next job, which is the same compounding loop that builds the Org's reputation over time.

Example

A bakery owner running two locations has no time left for Instagram, Google reviews, loyalty emails, or supplier follow-ups. She spins up an Org and switches on Marketing and Support. The Marketing agent plans campaigns and drafts content on the calendar; the Support agent replies to reviews and DMs and answers order questions around the clock. She reviews the drafts before they post and overrides the occasional support reply, while she stays focused on baking. Every approval and correction she makes is written back into the Org's memory, so both departments get sharper with each week that passes.

What to expect

With a department active, you should see the agent producing real output on its own, a filled content calendar, a moving pipeline, answered tickets, live community posts, while you spend your time on approvals and edits rather than production. Over repeated cycles, the Org's Neutron memory accumulates evidence from every job, and the department's agent needs fewer corrections from you as it goes.

Common mistakes

• Treating an app as full autopilot and never reviewing what the agent ships, which skips the approval step the apps are built around.

• Ignoring the option to take the wheel on a decision that matters, then being surprised by an agent's output later.

• Assuming each department learns in isolation, when in fact your edits and approvals feed the same Neutron memory every other department reads from.

• Expecting a department to perform perfectly on day one, before it has accumulated enough evidence in memory from real jobs and your corrections.

Related features

To see how these departments fit alongside an Org's identity, workforce, and treasury, read Anatomy of an Org. To understand how the evidence from every approval and correction turns into the memory these apps draw on and Dreaming works from, read Org Memory.