Written October 2026.
The pattern is familiar. The business needs something. Nobody on the team has done it before. It has to happen quickly, and the team has already made commitments that cannot slip. Someone has to dig in and figure it out, and the person with the most experience and the least slack in their calendar is usually me.
For a long time I took that as the job. I would dig in, work it out, and hand back a solution.
What I used to believe
The reasoning was sound on its face. The cost of the team stopping to learn something new was real, and the cost of me doing it was smaller. I could get there faster, and nothing else would be disturbed.
I have delivered many solutions that way, and I am proud of the ones where it changed an outcome. I am not going to pretend those did not matter. Some of them did.
But the pattern had a cost I could not see from the inside, because the cost never arrived on the day I solved the problem. It arrived later.
What challenged it
What changed my mind was noticing what came next. Every time I took one of these on myself, the team ended up more dependent on me, not less. The solution existed, but the understanding sat in one head, and it was not theirs. The next unfamiliar problem came to me again.
That is when I started reading the urgency differently. A problem that lands on the most senior person because nobody else can take it is often a symptom. Something upstream is missing: the team has not had the chance to build that kind of ground, or the environment is not giving them room to. If I solve the problem and leave that alone, I have treated the symptom and kept it comfortable.
I should be plain about my own part in it. Nobody asked me to be the single point of dependence. I built that, one reasonable decision at a time, and each decision felt like responsibility. It took me too long to see that it was also a way of avoiding the harder work of developing other people.
At this stage in my career, almost every chance to jump in and resolve a problem is not a chance for heroics. It is a chance for the team to level up.
How my thinking changed
The value I bring is to guide and mentor the people on the team, and to make sure they get the opportunities they deserve to grow from the environment around them. That is a different job from being the best solver in the room.
It comes down to this: knowing you can do something does not mean you should. In many cases the better investment is letting the team find the solution, because then they own the foundation they need to apply it again. In my experience, an answer I hand over tends to get used once. An answer a team works out tends to stay with them.
None of this is a case against being hands on. Giving up hands-on building is often the hardest part of the transition, and I have not given it up entirely. I still experiment with new technology and new ways to solve problems we already have. What changed is what I do with the result. The job is now to influence from that experience, rather than to ship the answer myself. I think that is one of the hardest moves most engineers make.
A plain way to decide
When a hard problem arrives and my instinct is to take it, I try to answer three questions, in order.
Is this an emergency that cannot wait for anyone to learn? If yes, I step in, and I say so out loud. I also say what I will do afterward so it does not happen the same way again. These cases are rarer than they feel while they are happening.
Is the team capable of getting there, with some support? Most of the time it is. Then I coach: I ask what they have tried, help them frame the problem, and stay available without taking the keyboard. It is slower, and usually it is the right call.
Does the same problem keep arriving? If it does, neither of the first two is the real answer. The question becomes why it keeps coming to me, and what in the way the team works, is organised, or learns would have to change. That is a design problem for the system, and it is where a leader’s time is best spent.
Take a hypothetical: a team keeps escalating the same kind of unfamiliar request. Solving each one quickly makes me look responsive. It also guarantees I will be asked again. Changing how the team gets time and cover to work on unfamiliar problems takes longer and is less visible, but it is usually the only one of the three that stops the escalations.
Why this is getting more pressing
AI tooling is accelerating this transition. As these tools take over more of the direct building, I expect the people who stay valuable to be those who can shape the outcome through judgment, context and the growth of the people around them, not only through what they personally produce. For engineers who built their identity on being the one who solves it, that is a real change, and I think it is arriving sooner than many of us expect.
It applies to me as well. I still build, including with these tools. But I try to treat what I learn as something to pass on rather than something to hold.
This week
Look back at the last problem you jumped into. Ask what the team learned from it, and whether the next similar problem would come to you again. If the honest answer is yes, pick one thing you can change so that it goes to someone else next time, and let them take it slowly.
If you have made this transition, or are in the middle of it, I would like to compare notes: connect on LinkedIn.