How AI is giving me back 5 hours every week
For the past few months, I’ve been introducing AI into my day-to-day work to help with the administrative tasks that come with being a Project Manager. We’ve all wished we had a little more time in the working week. And I’m sure we’ve all wondered how useful it would be to have our own personal assistant (PA) to take care of the smaller jobs.
So, what if AI could do just that?
I want to talk about the parts of the job that don’t usually make it onto your CV. The CRM admin, the inbox clean-up and all those little “I’ll update that record later” tasks that somehow pile up until Friday afternoon.
If you work in project management, you’ll know how much time goes into keeping your CRM and work-management systems up to date. Opportunities, quotes, project items, notes and deployments all need attention. They’re important, but they’re also supporting tasks that take time away from the work that really needs you.
At enable.services, we’ve been teaching our AI how we work. Not just giving it a general instruction to be helpful, but equipping it with specific skills and processes that reflect the administrative tasks we regularly complete in our CRM.
Here’s how it’s making a difference to my week.
Starting the day without the guesswork
My mornings used to start with multiple tabs open. Emails, the CRM, my calendar and a growing list of people I needed to get back to.
Now, I can simply ask our AI: “What should I do today?”
It looks across my projects in the CRM and gives me an overview of what needs my attention. That includes replies I owe, projects approaching their end dates, upcoming meetings I need to prepare for and newly raised feature requests.
Instead of spending the first part of my morning piecing everything together, I have a clearer picture of my priorities and can get straight into the work.
Keeping projects moving
One of my favourite features is how AI helps me keep track of everything that needs to happen during a deployment.
It’s easy for smaller requirements to get overlooked as a project progresses. For example, we might implement a layout change or a particular feature at the beginning of a project. It’s been logged and discussed, but by the time we’re approaching deployment, it can be easy to lose sight of it.
AI helps bring those details back to the surface, so nothing gets missed.
When items are ready to go live, it can also draft the deployment plan. This includes risks, mitigations, required steps, who’s involved and the checks outlined in our internal procedures.
I still review every word before anything progresses, but starting with a structured draft rather than a blank page makes a real difference. Our deployment checklists have never been more thorough.
Creating customer testing guides without the hours of admin
This is probably the biggest time-saver for me.
Whenever we complete a piece of project work, we provide the customer with a testing guide. This takes them through each change step by step, helping them check that everything works as expected.
For larger projects, these guides can run to eight or nine pages, covering different testing scenarios and checking how changes interact with existing functionality.
Previously, putting one together could take me two to three hours. I’d have project tasks open in one window, the original requirements in another, and spend time piecing together everything that had been implemented.
Now, AI does much of that groundwork.
It reads the project and relevant items in our CRM, identifies what’s been implemented, considers the impact on the existing system and produces a testing guide in our wiki using our established format.
My role is to review it, refine the wording and make sure everything makes sense from the customer’s perspective.
Reviewing a draft is a very different experience from starting with a blank page. It gives me more opportunity to focus on the details, check the testing scenarios and make sure the customer has everything they need, rather than spending hours assembling the document.
The honest bit: AI still needs managing
I don’t want to make this sound like a magic solution, because it isn’t. AI is only as good as the information it has access to and the instructions we’ve given it.
That’s why we’ve invested time in creating specific AI skills that explain how we work.
Most organisations have their own CRM quirks. Information might be stored in fields with labels that don’t quite match what you’d expect. If AI misunderstands where something belongs, a record might look complete without the information actually being accurate.
Teaching it our processes, terminology and rules is what makes it useful and trustworthy.
I’m still the one making the decisions. AI doesn’t independently assign tasks or move a deployment forward because it thinks a step is complete. It prepares drafts and recommendations for me to review and approve.
Anything customer-facing gets checked, and nothing progresses without the appropriate human input. We’ve also made sure our approach keeps customer data safe by running everything within our own systems.
