The Company Brain Launch Kit: Templates, Folder Structures, and Setup Playbook

When an organization stores its operational knowledge across disconnected chat threads, local hard drives, and fragmented cloud storage folders, human teammates and autonomous AI agents face severe context fragmentation. A developer prompting an AI coding assistant to refactor an internal API endpoint receives obsolete code because the architectural decision record lives inside an abandoned personal drive, while a client support lead answers an urgent account inquiry with superseded pricing because current contract terms sit in someone's private email thread. Traditional corporate wikis and dedicated enterprise search portals struggle to solve this problem because they treat knowledge maintenance as an extra chore separate from where daily work happens. Employees must manually format wiki pages, or engineering teams must construct custom vector databases with fragile chunking, embedding, and indexing pipelines that require ongoing maintenance.

A company brain launch kit is a curated set of folder taxonomies, permission baselines, connector setups, and team guidelines designed to take an organization from raw disparate files to an intelligent searchable brain with about ten minutes of wiring, after which syncing and indexing existing files continue in the background.

At its core, a company brain is a centralized operational environment where the documents a business runs on (contracts, specifications, architectural decisions, standard operating procedures, onboarding materials, meeting notes, and agent outputs) live together, index upon arrival, and answer natural language questions from both people and AI agents. Instead of forcing teams to learn specialized database querying tools or constantly maintain static documentation pages, an intelligent company brain grounds every inquiry in current primary source documents.

Dedicated enterprise search products, wiki tools, and legacy cloud drives each address isolated pieces of this challenge, but their architectural assumptions introduce distinct friction points:

• Enterprise Search Systems: Enterprise search systems like Glean index content across dozens of third-party apps and expose it to agents through MCP, but the documents still live in those apps, so there is no shared file workspace where agents save outputs beside the sources.
• Internal Knowledge Wikis: Notion AI and Guru can search connected apps such as Slack and Google Drive, but their own knowledge lives in pages and cards that people write and keep current by hand.
• Commodity Cloud Drives: Cloud drives like Google Drive, Dropbox, Box, and OneDrive store files reliably, and several now add AI assistants or MCP servers, but each covers only its own drive, so a team with files in several of them still has no single place where people and agents search and write together.

Fast.io bridges the gap between active file storage and artificial intelligence. In Fast.io, the files themselves live directly in the intelligent workspace where people and AI agents collaborate. When documents arrive through upload or cloud synchronization, the workspace indexes them on arrival for hybrid search (combining exact keyword matching and conceptual semantic retrieval). Both humans and AI agents can query the workspace, receiving direct answers that cite the specific files, page numbers, and snippet passages where the information originated.

• Knowledge Base Best Practices for Human Teams and AI AgentsKnowledge base best practices are operational standards for structuring, synchronizing, and securing central...
• 18 agentic workflows to run against your company brainYour agent starts every session knowing nothing about your company. A company brain fixes that. One Fastio workspace...
• How to Build a Company Brain for Customer Success TeamsA company brain for customer success unifies client contracts, onboarding records, and product documentation into an...
• How to Build a Company Brain for Finance Teams: Security, Isolation, and AI RetrievalA company brain for finance teams centralizes accounting policies, vendor agreements, audit files, and financial models...
• How to Build a Company Brain for Legal Teams: In-House Knowledge ManagementA company brain for legal teams centralizes executed agreements, standard clause playbooks, and corporate compliance...
• Building a Company Brain for Marketing Teams: Brand Assets, Messaging, and AI GroundingA company brain for marketing teams unifies brand guidelines, messaging frameworks, customer evidence, and competitor...

More on this subject: Agent Memory and Storage (335 guides)

Deploying an effective company brain begins with an intentional workspace architecture. A frequent mistake organizations make when setting up a company brain is treating the entire company as a single unstructured dumping ground. Without clear departmental boundaries, search queries can return irrelevant drafts or expose internal operational details to external guests.

Understanding Fast.io's access model is critical when designing folder taxonomies. In Fast.io, permissions are configured per organization, workspace, and share through member roles. Fast.io does not offer folder-level or file-level permissions within a workspace: every participant in a workspace accesses all files in that workspace according to their assigned role. Because workspaces do not enforce folder-level permissions, material with different audience requirements must live in dedicated workspaces or dedicated shares.

To establish clean boundaries, organizations should deploy four core workspace types, accompanied by branded shares for external delivery:

1. Operations and Company-Wide Knowledge Workspace

This workspace serves as the public square of the company brain. Every member of the organization belongs to this workspace to access shared company policies, operational guidelines, and organizational charts.

Company-Wide Knowledge/
├── 01_Company_Overview/
│   ├── Mission_and_Strategy.md
│   ├── Executive_Bios.pdf
│   └── Organizational_Chart.pdf
├── 02_Policies_and_Handbooks/
│   ├── Employee_Handbook_2026.pdf
│   ├── Travel_and_Expense_Policy.pdf
│   └── Remote_Work_Guidelines.md
├── 03_Standard_Operating_Procedures/
│   ├── IT_Helpdesk_and_Hardware_Requests.md
│   ├── Information_Security_Baselines.md
│   └── Facility_and_Office_Guidelines.md
├── 04_Brand_and_Marketing_Assets/
│   ├── Brand_Style_Guide.pdf
│   ├── Official_Logos_and_Typography/
│   └── Presentation_Templates.pptx
└── 05_Meeting_Memos/
    ├── All_Hands_Presentations/
    └── Quarterly_Company_Updates/

2. Product and Engineering Workspace

Designed for technical teams, software engineers, and autonomous coding agents, this workspace holds the architectural truth of your product systems. Coding agents connected via MCP inspect these folders to ground code changes in verified system architectures.

Product and Engineering/
├── 01_Architecture_Decision_Records/
│   ├── ADR-001_Data_Persistence_Strategy.md
│   ├── ADR-002_Authentication_and_Session_Handling.md
│   └── ADR-003_API_Gateway_Specifications.md
├── 02_Product_Specifications/
│   ├── PRD_Core_Search_Engine_2026.md
│   └── Scope_Mobile_Workspace_Sync.md
├── 03_API_Documentation/
│   ├── OpenAPI_Public_Spec_v2.json
│   └── Webhook_Payload_Definitions.md
├── 04_Infrastructure_Runbooks/
│   ├── Disaster_Recovery_Playbook.md
│   ├── Database_Migration_Checklists.md
│   └── Incident_Severity_Definitions.md
└── 05_Release_Notes/
    └── Changelogs_2026.md

3. Go-To-Market Workspace

The revenue engine requires immediate access to battle cards, customer case studies, and renewal tracking. Sales representatives and AI research agents consult this workspace to prepare meeting briefs and respond to customer requests for proposals.

Go-To-Market/
├── 01_Sales_Playbooks/
│   ├── Enterprise_Pitch_Deck_2026.pptx
│   ├── Objection_Handling_Matrix.md
│   └── Demo_Environment_Runbook.md
├── 02_Competitive_Intelligence/
│   ├── Competitor_Displacement_Battle_Cards.md
│   └── Market_Pricing_Analysis.xlsx
├── 03_Customer_Case_Studies/
│   ├── Manufacturing_Logistics_Case_Study.pdf
│   └── Global_Financial_Services_Deployment.pdf
├── 04_Renewal_Briefs/
│   ├── Tier_1_Account_Health_Summaries.md
│   └── Annual_Contract_Schedules.pdf
└── 05_Win_Loss_Analyses/
    └── Q3_Enterprise_Win_Loss_Review.md

4. Restricted Confidential Workspace

Certain enterprise assets must remain strictly isolated from the broader employee base and unauthorized AI agents. Executive compensation, board presentations, audited financials, and active legal disputes reside in this workspace.

Executive and Legal Confidential/
├── 01_Board_and_Governance/
│   ├── 2026_Board_Meeting_Minutes.pdf
│   └── Investor_Reporting_Packets/
├── 02_Financial_Models/
│   ├── Annual_Operating_Budget_2026.xlsx
│   └── Cash_Flow_Projections.xlsx
├── 03_Legal_and_Compliance/
│   ├── Master_Services_Agreements/
│   └── Regulatory_Filing_Disclosures.pdf
└── 04_Executive_Compensation/
    └── Equity_Incentive_Plan_Allocations.xlsx

5. External Delivery and Client Portals

When sharing materials with external clients, auditors, or legal counsel, teams should not invite external parties directly into internal workspaces. Instead, teams create branded shares (Send, Receive, or Exchange) pointing at specific folders. External recipients open shares through secure web links without creating an account or consuming an organization seat license.

Standardizing workspace taxonomies reduces document retrieval friction across remote teams by providing predictable file locations and clear retrieval paths. Per-file version history tracks every modification automatically. When teammates or autonomous agents update documents, previous versions remain intact and can be restored at any point.

A company brain contains proprietary intellectual property, strategic roadmaps, and sensitive client documentation. Establishing strict access control baselines prevents unauthorized internal discovery and safeguards data boundaries before connecting AI assistants.

Fast.io structures access across two administrative tiers: organization roles and workspace roles. Understanding the boundary between these tiers ensures that sensitive company knowledge remains restricted to appropriate members:

• Organization Roles: Applied across the entire subscription account. The Owner maintains billing, plan administration, and organization-level security.

Admins can create new workspaces, invite team members, manage integrations, and review administrative settings.

Members can participate in workspaces to which they have been granted access.

• Workspace Roles: Configured individually within each distinct workspace. Roles include Owner, Admin, Member, Guest, and View. A member assigned the View role can inspect files and run queries against workspace intelligence but cannot upload, edit, move, or delete documents.

Ownership transfer provides operational flexibility for administrative handoffs. An organization owner or workspace owner can transfer ownership to another team member. When ownership is transferred, the previous owner is automatically demoted to admin. This pattern is particularly useful when an external consultant or automated setup agent configures an organization and hands primary ownership to a corporate executive.

A common governance risk in growing organizations is unrestricted workspace access. In Fast.io, a workspace's join setting defaults to letting any member of its organization join. While this open join setting works well for general company handbooks and brand assets, administrators must restrict the default workspace join setting on confidential workspaces, such as Executive, Finance, and Legal. With the setting restricted, people must be added to the workspace directly instead of joining on their own, which prevents accidental cross-departmental data exposure.

Fast.io Access Control Hierarchy:

Organization (Owner, Admin, Member)
  ├── Company-Wide Knowledge Workspace (Join Setting: Open to Org)
  │     ├── Role: Member (Full search and document access)
  │     └── Role: View (Read-only search and file inspection)
  │
  ├── Product & Engineering Workspace (Join Setting: Open to Org)
  │     ├── Role: Member (Developers, Technical Writers)
  │     └── Role: Admin (Engineering Managers)
  │
  └── Confidential Executive Workspace (Join Setting: Restricted)
        ├── Explicitly Invited: Executive Leadership
        ├── Explicitly Invited: General Counsel
        └── Uninvited Organization Members: Zero Access

When collaborating with outside contractors, vendors, or legal advisors, administrators should deploy share-level controls rather than granting full workspace access:

• Branded Send Shares: Distribute finalized reports, investor decks, or sales collateral. Administrators can disable file downloads so external recipients review documents in the browser instead of saving copies.
• Branded Receive Shares: Collect audit records, vendor invoices, or compliance documents from outside parties. External partners upload files directly to a designated workspace folder through a web portal without viewing existing folder contents.
• Branded Exchange Shares: Establish a two-way collaborative portal where both parties upload and review documents for a specific initiative.

Security controls also govern AI agents and programmatic integrations. In Fast.io, API keys carry explicit scopes and act strictly with the permissions of the user account that created them. API keys have no independent role of their own. When configuring an MCP connector with a Read Only scope, tools that modify data are not offered to the model at all. Because data-modifying tools are excluded from the tool schema, the agent cannot attempt unauthorized file edits, deletions, or permission alterations.

All workspace modifications are tracked within a detailed activity log. The activity log records changes across the organization, capturing uploads, edits, moves, deletions, membership adjustments, permission changes, and share creations. This provides an audit trail of how files and structures evolve over time.

Most organizations already maintain substantial document libraries in existing cloud drives. Moving that information into an intelligent company brain should not require manual downloads, local drive juggling, or complex API migration scripts.

Fast.io provides built-in Cloud Sync to bridge historical file repositories into modern intelligent workspaces. Cloud Sync connects Google Drive, Dropbox, Box, and OneDrive folders directly to a workspace. For organizations operating on Microsoft 365, SharePoint document libraries connect directly through the OneDrive connector.

Synchronization can run one-way or two-way, on a schedule or on demand. It is never continuous, live, or real-time, so teams choose when changes are pulled in. Teams can also run one-time Cloud Imports from Google Drive, Dropbox, Box, or OneDrive, or import a file from a public URL, moving document archives between cloud providers without consuming local workstation bandwidth.

A fundamental security principle governs all imported data: files synced or imported from another drive take the Fast.io workspace's permissions, not the source drive's sharing settings. If a legacy Google Drive or SharePoint folder had loose, outdated sharing links across past contractors, syncing that folder into a restricted Fast.io workspace confines those documents strictly to authorized participants in that Fast.io workspace.

Once documents land in a workspace with Intelligence enabled, Fast.io automatically indexes them for hybrid search without requiring external vector databases or third-party retrieval infrastructure. The intelligence engine combines full-text lexical indexing with dense semantic vector representations:

• Lexical Full-Text Search: Quickly locates exact part numbers, software error codes, employee names, and literal legal clause citations.
• Semantic Vector Retrieval: Understands conceptual meaning and synonyms, matching inquiries like "health insurance deductible limits" to passages discussing "out-of-pocket medical maximums."
• Citation-Backed Responses: When employees or AI agents query the workspace, answers include direct citations referencing the specific document, page number, and text excerpt used to formulate the reply.

Beyond conversational search and summaries, organizations often need to extract structured data from unstructured documents. Fast.io provides Metadata Views to turn documents into live, queryable databases. Rather than building custom optical character recognition (OCR) scripts, users describe the fields they want in natural language. Fast.io generates a typed schema supporting Text, Integer, Decimal, Boolean, URL, JSON, and Date & Time fields, automatically extracting structured values from PDFs, presentations, and spreadsheets into a sortable, filterable table.

Example: Metadata Views Contract Registry Schema

Document Name          Counterparty      Effective Date    Renewal Term    Governing Law    Auto-Renew
------------------------------------------------------------------------------------------------------
Vendor_Agreement_A.pdf Acme Software     2026-01-15        12 Months       Delaware         True
Facility_Lease_HQ.pdf  Prime Properties  2025-06-01        36 Months       California       False
SaaS_Order_Form_04.pdf DataStream Inc    2026-03-20        24 Months       New York         True

Teams deploy Metadata Views across departments: legal teams track contract renewal dates and counterparties, finance teams extract invoice totals and payment terms, and engineering teams catalog API specification versions across technical whitepapers.

Deploy Your Company Brain

Set up structured workspaces, connect team files, and configure MCP agents for intelligent search with source citations. Start with a 14-day free trial on any monthly plan.

Start 14-Day Trial

A modern company brain serves two distinct audiences: human team members reading documents in web interfaces, and autonomous AI agents analyzing files, answering operational questions, and generating new work.

Connecting AI agents to enterprise storage has historically required brittle custom scripts or complex integration code. Fast.io solves this by offering native support for the Model Context Protocol (MCP), an open standard designed to connect AI applications directly to external data sources and tools.

Anthropic documented the foundational vision of the protocol upon its public release:

The Model Context Protocol is an open standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools.

Fast.io provides a remote MCP server over Streamable HTTP. It requires no local npm installations, background daemon processes, or intermediate bridge utilities, and it has a separate address for each kind of client:

• Interactive Assistants and Personal Agents: Applications such as Claude (web, desktop, and mobile), OpenClaw, and Hermes Agent connect to https://mcp.fast.io/mcp/tools. Users connect via OAuth in their browser, selecting whether the connection is Read Only or Read & Write and specifying which organizations and workspaces the agent can reach. Complete setup documentation is available at mcp.fast.io/docs, with agent skill specifications at mcp.fast.io/skill.md.
• Coding Agents and IDE Environments: Development assistants including Claude Code, Cursor, VS Code with GitHub Copilot agent mode, Cline, and Devin connect directly to https://mcp.fast.io/mcp/code.
• ChatGPT and OpenAI Codex: Users install the Fastio plugin directly from the ChatGPT plugin directory, or configure a custom MCP server pointing to https://mcp.fast.io/mcp/operations.

Once the connection works, give the agent an Ask the Brain skill: an instruction to query the company brain workspace before it drafts, edits, plans or triages anything, to ask follow-up questions until it has the whole picture, and to cite the file behind each factual claim. Then add one line to the agent's memory telling it to use that skill whenever a task needs company context.

To protect workspace integrity, Fast.io's MCP endpoints enforce a strict architectural separation between read actions and write actions. On https://mcp.fast.io/mcp/tools, inspection queries execute through the storage and share tools, while file creations, moves, and deletions require calls to the corresponding storage_manage and share_manage tools. When an agent connects with Read Only permissions, the _manage tools are completely absent from the tool schema, ensuring the agent cannot execute write operations.

Fast.io also includes Collaborative Notes kept beside files in the workspace. Notes are markdown documents that people and agents can edit together. Collaborative Notes support live multiplayer co-editing, displaying visible labeled cursors for both human teammates and AI agents.

To coordinate edits during multi-agent collaboration, Fast.io provides advisory Agent Intents. An AI agent or human teammate can declare an intent to announce the specific note or document they are actively updating. Agent Intents are advisory and do not lock files or block concurrent writes, allowing teams to coordinate smoothly while version history records every iteration.

// Example: Cursor configuration in ~/.cursor/mcp.json
{
  "mcpServers": {
    "fast-io": {
      "url": "https://mcp.fast.io/mcp/code"
    }
  }
}

A common workflow pattern involves a research agent querying the company brain via MCP to synthesize recent product specifications and competitive battle cards, drafting a new product requirement note. A human product manager opens the Collaborative Note, sees the agent's active cursor completing the draft, and immediately begins refining the text beside the agent.

Transforming fragmented company files into an active, intelligent company brain does not require months of consulting or engineering overhead. Following this four-step checklist, the core wiring takes about ten minutes; syncing and indexing existing files then continue in the background, and how long they take depends on volume:

1. Workspace Structure and Join Permissions

• Sign up for Fast.io and establish your primary organization.
• Create your baseline workspaces: Company-Wide Knowledge, Product and Engineering, Go-To-Market, and Executive Confidential.
• Open workspace settings for Executive Confidential and restrict the default join setting from open to restricted, ensuring that only explicitly invited members can enter.
• Invite team department leads with appropriate workspace roles (Member or Admin).

2. Cloud Sync Connector Setup

• In the Company-Wide Knowledge workspace, open Cloud Sync and select your existing cloud storage provider (Google Drive, Dropbox, Box, or OneDrive).
• Authorize the connection and select the source folder holding your official policies, employee handbooks, and operational SOPs.
• Configure a scheduled one-way sync to regularly pull updated documentation into the company brain.
• Confirm that files begin indexing as they arrive. A large folder keeps syncing and indexing in the background after the wiring is done, and its files become searchable, with snippet citations, as each one is indexed.

3. MCP Agent Configuration

• Connect your team's preferred AI desktop applications and coding tools.
• In Claude Desktop, open Customize, select Connectors, add a custom connector, and enter https://mcp.fast.io/mcp/tools. Sign in through OAuth and grant access to your designated workspaces.
• In Cursor or VS Code, add the Fast.io MCP endpoint https://mcp.fast.io/mcp/code to your project configuration.
• Test agent connectivity by asking the AI to summarize an internal policy or locate a specific architecture decision record.

4. Team Prompt and Search Handbook

• Share a practical prompt reference handbook with team members, demonstrating how to query the company brain using natural language:
  • Operational Inquiries: "What is our official travel reimbursement policy for domestic conferences?"
  • Meeting Preparation: "Summarize the key objections raised in our last three enterprise renewal briefs."
  • Technical Grounding: "Find the architecture decision record explaining why we chose our current session authentication model."
  • Onboarding Orientation: "Where are the setup runbooks for new engineering workstation provisioning?"
• Instruct employees to click through on citations to inspect the primary source pages whenever verifying critical facts.

Operational governance remains simple following rollout. Administrators monitor workspace modifications through the detailed activity log, which captures all file uploads, edits, moves, deletions, and membership modifications made by people and agents.

Fast.io offers transparent pricing designed for growing organizations. Monthly plans start with a 14-day free trial, which requires a credit card. Paid subscriptions scale across three tiers:

Credits meter AI work only. Storage and seats come with each subscription plan. Resource overages and add-on pricing follow transparent rates:

By providing an intelligent foundation where documents live, index, and co-exist with autonomous agents, Fast.io transforms fragmented company knowledge into an authoritative company brain.