More present with AI

| 3 min read

One thing has bothered me for a while about engineering management: too much of the time that should go into people goes into collecting information about them.

Before a 1:1, I want to remember what we discussed last time, what’s happened in their team since, and where they might need my help. When I talk to my EMs, I want to already know what’s moving and what’s stuck.

Being prepared is, to me, a form of respect for people’s time and their work. But it requires a surprising amount of machine-like work: collecting updates, finding old conversations, remembering actions, preparing context.

So over the last few weeks, I’ve been building daily routines on top of AI. Not a chatbot I ask questions to, but automations that prepare my day and post it on Slack before I open my laptop:

  • A “secretary” that preps my 1:1s from previous notes and actions.
  • An interview-prep routine that pulls a candidate’s CV from my inbox and previous test results.
  • A daily Jira health check across my teams’ initiatives and epics, compared with the day before, so I can see what’s moving and what’s stuck before talking to my EMs.
  • An end-of-day pass that turns meeting notes into a local knowledge base, cross-referencing conversations, objectives and interactions across my teams and stakeholders.

That last one should also mean end-of-year reviews are no longer a scramble through six months of scattered notes and Slack messages. The context is already building itself, week by week.

And the whole thing uses roughly $0.30 a day of my Claude quota.

But saving time isn’t really the point.

I want machines to do the machine-like part of my job, so I have more time and headspace for the human part. I want to spend a 1:1 understanding where someone needs help, what is frustrating them and where they want to grow. I want conversations with my EMs to be about making better decisions and helping them develop, not asking them to repeat information I could have gathered beforehand.

AI isn’t replacing those conversations. It’s allowing me to be more prepared, more knowledgeable about people’s work and, ultimately, more present with them. None of this replaces judgement. It removes the hunting and gathering that used to eat the time and attention that judgement, and people, deserve.

I can’t share these directly: they’re wired into my own context and workflow. But I’m turning the useful parts into neutral, easy-to-install and adaptable versions to share with my colleagues. And with a bit of creativity, this post and Claude, anyone could build the same kind of system around the way they work.

What’s the one machine-like part of your job you’d happily give back to a machine?

Originally published as a LinkedIn post.