All articles

Aiver Ilaya
Aiver IlayaDirector, Aliso Digital
ShareLinkedIn
Aiver Ilaya, director of Aliso Digital

Written by

Aiver Ilaya.

Director of Aliso Digital, a Melbourne-based IT consultancy specialising in digital transformation programs, process mapping and documentation.

IT documentation has always had 2 halves.

There’s the mechanical half.

Reading through 60 legacy documents, formatting release notes, chasing terminology inconsistencies, restructuring things into the current template.

Then there’s the half that decides whether the documentation works.

Sitting with an SME until you understand what actually happens, not what the procedure claims happens.

The mechanical half used to eat most of the week.

That’s the part that’s changed, so the interesting question is what you do with the difference.

📚 Reviewing existing documentation

Every IT program starts with an inherited pile.

Procedures from a system that’s been upgraded twice.

Guides written by 4 different people with 4 different structures.

A knowledge base where nobody’s certain which articles are still accurate.

Reviewing that properly used to take days of reading before you’d written a single word.

Now you can put the set in front of a tool, then ask specific questions.

  • Which of these documents contradict each other?
  • Where does this guide reference something that doesn’t exist elsewhere?
  • Which procedures mention screens or menus that changed in the last release?
  • What’s covered twice, in different words, across separate documents?
  • What would a new starter still not know after reading all of it?

The answers need verifying, obviously.

Still, going from a rough picture in an hour rather than 3 days changes what’s realistic when a program wants a documentation plan by Friday.

More usefully, it tells you where to spend your reading time rather than reading everything evenly.

📝 Release notes

Release notes are a good example of work that’s important, repetitive, thankless.

Every sprint, someone has to turn a list of tickets into something a business audience can read.

The raw material is already there, sitting in the change records, the ticket titles, the developer comments.

The work is translation, not invention.

That’s exactly the kind of task worth handing over.

  • Grouping changes by what the user actually experiences, not by component
  • Converting ticket language into plain description
  • Separating what needs action from what’s informational
  • Flagging changes that affect existing procedures
  • Producing a short version for a broadcast email, a longer one for the knowledge base

What still needs a person is judgement about impact.

A developer’s minor fix can be a significant change for a team of 200 people, since the field they rely on has moved.

Only someone who knows the users spots that.

⏱️ Turnaround time

Shorter turnaround matters more than it sounds on an IT program.

The old cycle was slow enough that documentation arrived after people needed it.

A system change lands, the procedure gets updated 3 weeks later, meanwhile the service desk fields questions it shouldn’t have to.

Faster drafting closes that gap.

A change goes in Tuesday, the updated work instruction can be out Wednesday, verified rather than rushed.

There’s a second effect worth mentioning.

When updates are quick, people stop treating documentation as a fixed thing they have to live with.

They report errors, because they know something will happen.

That feedback loop is worth a lot, though it only starts once turnaround is short enough to be credible.

🤝 The point of all of it, more SME time

Here’s what actually matters.

The constraint on IT documentation quality was never writing speed.

It was access to the people who know things.

An engineer with 40 meetings a week.

A service desk lead mid incident.

A finance manager who’s been asked the same questions by 2 other projects this year.

Time saved on mechanical work goes straight into that relationship, if you let it.

What that looks like in practice.

  • Turning up already knowing how the platform works, so you’re asking them to correct you rather than educate you
  • Booking 20 minutes instead of requesting an hour, because you’ve done the preparation
  • Sending a draft within a day, while the conversation is still fresh in their mind
  • Following up with the quiet person from the workshop, who usually holds the exception handling
  • Having capacity for a second conversation, rather than making the first one carry everything

That last point changes the quality of a documentation set noticeably.

The first conversation gets you the standard path.

The second gets you what happens when it goes wrong, which is the part users actually need written down.

🔍 The trade you’re making

None of this works if the speed gets treated as an excuse to check less.

The load doesn’t disappear, it shifts from producing to verifying.

Verifying is more demanding than transcribing, so keep the habits tight.

  • Read every draft line by line
  • Test each step against the live system
  • Keep the SME review, since a clean draft still needs a human check
  • Watch for the moment you accept output because you’re tired, not because it’s right

Fatigue makes plausible text look correct, which is the main risk in all of this.

✅ Where it lands

AI hasn’t changed what good IT documentation is.

It’s changed how much of your week goes into producing it versus understanding the thing you’re documenting.

More collaboration, less formatting.

That’s a straightforwardly better trade, provided the verification stays in place.


About the author

For 19 years Aiver has worked on enterprise digital transformation programs – multi-million dollar programs running years at a time, for national organisations and global firms, reaching millions of end users.

An IT consultant based in Melbourne, his experience spans ITSM, HRIS, ERP, CRM, POS, KMS and EDRMS platforms, plus workflow automation, covering business analysis, process optimisation, technical writing, knowledge management and UI / UX design.

He has worked across government, healthcare, science and medical research, finance, education, telecommunications, infrastructure, aviation, construction and property, technology, retail, logistics, pharmacy, insurance and professional services.

He has run his own consultancy for 12 years and works with a small number of clients at a time. Available for remote and hybrid contracts.

alisodigital.com

Read More

Related Posts

AI Does the Boring Half of IT Documentation, Which Leaves Room for the Useful Half

IT documentation has always had 2 halves. There’s the mechanical half. Reading through 60 legacy documents, formatting release notes, chasing terminology inconsistencies, restructuring things into the current template. Then there’s the half that decides whether the documentation works. Sitting with an SME until you understand what actually happens, not what

BPMN 2.0 Is the Easy Part of Process Mapping

You can learn BPMN 2.0 properly in about a fortnight. Tasks, sequence flows, exclusive gateways, parallel gateways, start events, message events, pools, lanes. Maybe 12 elements cover most real world maps. The notation is well documented, widely understood, consistent from one organisation to the next. Which means it isn’t what

The First 14 Days After Go Live Decide Whether Your Rollout Sticks

Go live day usually feels like the finish line.The launch email goes out.The project team takes a breath.Then the tickets start. The first 2 weeks after go live are where a rollout either settles in or starts to slip.That period has a name.It’s called hypercare.Most teams plan for go live,

The Quiet Negotiation Behind Every Document Review

Nobody tells you this at the start of a technical writing career. Most of the job isn’t writing. It’s getting 20 minutes from someone who doesn’t have 20 minutes. You’ve drafted the procedure, you’ve mapped the process, you’ve done the hard part. Then it sits with a subject matter expert