Services

Most organisations arrive with one of these and discover the other on the way. A file estate nobody can account for and a process still being fed by scanning and rekeying are usually the same problem at different stages: information in a form the organisation cannot use.

Knowing what you hold

Shared drives accumulate for decades. Staff leave, departments reorganise, a migration gets half finished, and the naming conventions that made sense to one team in 2011 are opaque to everyone now. The result is an estate nobody can describe and nobody feels able to touch.

The first engagement is almost always an assessment. A read-only scan of the locations you nominate, producing a picture of what is actually there:

• Volume by location, and how that has grown
• Age profile — when files were last modified, not when they were created
• Duplication, including near-duplicates and superseded versions
• Orphaned content, where the owner has left and nobody inherited it
• Permissions that have drifted from what anyone intended

Nothing is moved, changed or deleted at this stage. The output is a document you can take to the people who have to authorise what happens next.

Disposing of it defensibly

Most organisations keep everything, because disposing of anything feels riskier than paying for another terabyte. That instinct is reasonable and it is also why the problem compounds.

The work is less about deletion than about being able to justify it afterwards. That means retention rules written down and agreed, legal hold applied where it belongs, and a record of what was removed, when, under which rule, and on whose authority. Disposal runs in stages, and each stage stays reversible until it is signed off.

Where an organisation has a statutory retention schedule, that becomes the spine of the rules. Where it does not, agreeing one is usually the most durable part of the engagement.

Making the remainder findable

Clearing the redundant majority is only half the value. What remains has to be easier to work with than what was there before, or the estate reverts within a year or two.

That generally means a sensible structure, consistent metadata on the material that matters, and search that returns the current version rather than eleven candidates. Where a SharePoint or cloud migration is planned, this is the work that should happen first — migrating an unassessed estate moves the problem rather than solving it.

Turning paper into process

The second line of work starts where documents enter the organisation: post, scanned attachments, forms submitted on paper, PDFs that are really just photographs of text.

Typically that material is read by a person and typed into a system before anything can happen. The cost is not only the typing. It is the queue in front of it, the errors that get introduced, and the fact that nothing downstream can start until someone has got to it.

The work involves:

• Capture and classification of inbound documents by type
• Extraction of the fields that matter, including from image-only PDFs
• Validation rules, so that what reaches your systems is checked rather than merely transcribed
• Exception handling, because a process that works for 95% of documents and silently mangles the rest is worse than no automation
• Integration with the systems that already hold the process

The measure is end-to-end time and error rate, not how much text was extracted.

How engagements are shaped

Most relationships start with an assessment, because it is the only honest basis for scoping anything larger. If the answer turns out to be that you have a smaller problem than you feared, that is a good outcome and a short engagement.

How an engagement runs sets out the sequence in more detail.