Repository

Every organization has one. Sometimes several. The colleague who has been there eleven years and knows where everything is, who commissioned what, why the 2021 pricing study was shelved, and which folder in which drive has the real version of the segmentation deck.

They are the most valuable knowledge infrastructure your company owns, and they are entirely undocumented, unreplicated, and available only during business hours. When they leave, the institutional memory of your organization leaves with them, and you find out about it gradually, in a series of small expensive surprises.

The Repository is the alternative. It is the foundation the rest of the platform sits on: one governed, searchable, permissioned home for what your organization knows.

What goes in

Almost everything. Reports, decks, PDFs, spreadsheets, Word documents, images, audio, video, interview transcripts, dashboards, web pages, and links to systems that live elsewhere.

Content arrives four ways.

Direct upload, including bulk. Drag in a folder, an archive, or a decade of accumulated research.

Integrations with the systems your teams already use. Box, Confluence, Dropbox, Microsoft Teams, ServiceNow, SharePoint, Slack, and Salesforce. Content flows in from where people already work rather than requiring them to change habit first.

Sharp It, our one-click browser capture for anything on the web. Covered in detail on its own page. It is how external content stops depending on somebody's bookmark discipline.

Migration, which we do for free, from whatever you are using now. That is not a promotional gesture with fine print. Migration is the single biggest reason knowledge platforms fail to launch, so we removed it as an obstacle entirely.

What happens on arrival

This is the part that separates a repository from a folder.

When content lands, our AI reads it. It generates a descriptive headline, a concise summary, and a set of tags. It identifies key themes. It connects the item to related material already in the hub. Audio and video are transcribed so they are searchable as text.

Nothing about this depends on the person uploading doing anything correctly. They do not need to name the file well, pick the right category, or write a description, which is fortunate, because they were never going to.

We are blunt about why this matters. Every knowledge management project in history has started with somebody designing a beautiful taxonomy and a tagging guide. Every one of them has ended the same way, which is with people uploading files named Final_v4_USE_THIS.pptx into whatever folder was open. Manual curation does not survive contact with a busy organization. Automated processing does, because it requires no discipline from anybody.

Structure: hubs, not folders

A hub is a scoped knowledge environment with its own content, membership, permissions, and branding.

Most organizations run several. A central insights hub. A hub per business unit, brand, or region. A competitive intelligence hub. A hub for an agency partner. A hub for a specific program or transformation.

Content can live in one hub or several. Permissions run from the hub level down to the individual item. This lets you build the structure your organization has, including the parts that are legally required.

Two examples from customers.

A bank maintains separate hubs for retail, commercial, and wealth, with an information barrier between certain functions that is a regulatory obligation rather than a preference, plus a shared economic research hub that everyone can reach.

A consumer goods company runs a hub per brand, with the shopper and category research shared across all of them, and a scoped agency hub containing exactly what the agency should see and nothing else.

Use cases, by the role that lives in here

Insights and research teams: the end of the archaeology expedition

Your studies, your vendor deliverables, your trackers, your syndicated subscriptions, and your raw materials in one searchable environment. The immediate effect is that the request "can you send me the thing about X" stops consuming your week.

The longer-term effect is more valuable. Three years of accumulated research becomes a body of knowledge rather than a pile of files, and you can answer questions about the body rather than about individual studies.

Competitive intelligence: a durable record instead of a personal archive

CI collection is usually somebody's private system of folders and bookmarks, and it evaporates when they change jobs. In a hub it becomes an organizational asset with history, so you can answer how a competitor's positioning has shifted over eighteen months rather than what they announced last Tuesday.

Knowledge managers and librarians: governance you can apply

Retention rules, archival, duplicate detection, ownership assignment, and content lifecycle management. You can find the 400 documents nobody has opened since 2021 and retire them, which improves search quality for everyone.

You also get the thing that makes the role sustainable, which is a system that does the classification work so your job becomes curation and program design rather than manual filing.

Compliance, risk, and legal: a defensible record

Audit trails, access logs, permission enforcement to the item level, and enterprise security including SOC 2, ISO 27001, GDPR compliance, AES-256 encryption at rest, and TLS 1.3 in transit. SSO and SCIM provisioning mean hub access follows employment status automatically rather than depending on somebody remembering to revoke it.

Onboarding and transitions: three months of context, available immediately

New hires, internal moves, and the brand manager rotation that happens every eighteen months. Instead of six weeks of meetings, they open the hub and read. Their manager can see through engagement analytics what they have covered and coach on the gaps.

Executives and chiefs of staff: a single source to point at

When two teams present conflicting numbers, the question is which analysis they came from. When both live in the repository with their sources and dates attached, that argument resolves in two minutes instead of two meetings.

Consultants and agencies: client work that stays organized

Deliverables, working materials, and prior engagements in client-scoped hubs. Entitlement boundaries between clients are enforced structurally rather than by everyone being careful.

What we are not

We are not a file storage system. If you need somewhere to keep 40 terabytes of raw video and every draft of every document, keep using your drive. The repository is for knowledge you want found and used, which is a smaller and more valuable set. Most customers connect their storage rather than replacing it.

We are not a wiki. Wikis are for collaboratively authored text and they are good at that. The repository holds the artifacts your organization produces and acquires, in the formats they arrive in, with the processing that makes them findable.

We are not a general enterprise search tool. Those index everything you own, which sounds appealing until you realize that everything you own includes years of low-signal material. We work over a curated environment, which is why results reflect deliberate collection rather than whatever happens to be lying around.

We are not going to pretend curation is unnecessary. Automated processing removes the tedious part. Deciding what belongs, what is authoritative, and what should be retired is still human work, and the customers who get the most from us have someone who owns that.

Objections we hear, answered

"We already have SharePoint." You have storage and you likely have a search box. What you do not have is automatic summarization and tagging, connection between related material, a distribution engine, or engagement reporting. Keep SharePoint. Connect it. Use it for what it is good at.

"Migration will take a year." We do it for free and we have done it from every major platform and from filesystems that had never been organized. The processing happens on arrival, which is why customers can search meaningfully in the first week rather than after a six month cleanup project.

"Our content is a mess." Everyone's content is a mess. That is the normal starting condition and it is the reason the processing pipeline exists. If your content were already well organized you would not be reading this page.

Try this

Pick a study your organization commissioned two or three years ago. Real budget, still relevant.

Ask someone who was not involved to find it, using only a plain language description of what they need. Give them ten minutes.

Most people try three filename guesses, run a keyword search, scroll through shared folders, and then message the person who might know. That last step is your actual knowledge management system, and it stops working the day that person changes jobs.