Inherited a project late in execution.
Overseas developer.
When I insisted on inspecting the work before the deadline, it was defective.
Escalated, cut the vendor and finished it myself.
Project succeeded, still in use years after I left.
What would you do in this situation?
Jensen Huang’s recent interview stuck with me: one of the best skills in the AI age is knowing how to ask good questions.
AI can give the answers and it’s on us to seek them.
I once inherited a project at the end of the Waterfall.
The deliverables met the requirements but I found out later it did not meet the users’ expectations.
They accepted the work anyway because they couldn’t be bothered with it anymore.
Project failed successfully.
It helps to pay attention to your surroundings when working on something sensitive.
If you need to work in public, choose a seat where no one can see your screen.
Do not say sensitive details out loud.
In the office, lock your computer before you step away.
Be paranoid.
That one time you hit a critical problem you’d never seen before.
You couldn’t quit or escalate it away.
But every attempt you made made it worse.
What did you do to solve it and what would you do differently now?
When you’re working on something critical with a deadline, start the backup plan at the same time, in case the main task fails or falls short.
Like starting the escalation while you’re still fixing a critical server. Waiting until you’ve failed only adds downtime.
@kunchenguid Censorship. People based in sanctioned countries will not have access, apart from the Chinese ones I guess. Also, if I want to train models on proprietary materials, some models do not allow that due to copyright.
When you are the error
User: Hey, we have a problem with…
Me: One moment, let me have a look…
Realizes I was causing it.
Me: Should be fixed now.
User: That was quick, thanks!
Perfect CSAT