Christopher Kouzios

Start small but think big. Consider what component of a larger enterprise program could benefit from AI and target a low risk sub component. You can iterate quickly but without spending a fortune or taking on huge risk. Isolate your use case, limit funding, prove your MVP -> build from there. Hard to argue (you can replace “AI” with any technology and the same principle applies).

Slow and steady wins the race if you’re budget constrained. Begin to migrate compute services as the hardware and/or software ends its life cycle. Stay away from database migrations as those can be pricey - wonderful but pricey. Some key ways to maximize your cloud budget: Tier 1 applications that need resiliency, virtual desktops (consider the amount of effort saved by not having to recycle laptops) and especially backups on low tier storage. Not the best options but some of the best ways to get started when on a budget.

Do your job. Period. Do not wait to be asked. Do not ask permission. Do not seek clarity before acting. Understand what your boss needs and fill that need. That’s the baseline. Then understand what they might need that they’re not sure they need but you see it and handle it. Be the solution to problems they know they have and the ones they do not yet see. When I was a little kid I worked for my dad who was a contractor. He used to say “why are you standing there watching me work? I don’t pay you to spectate. Go do something. Go clean something. If you don’t have anything left to do, clean the truck.” He had the cleanest most well organized truck in the history of contracting for years. I learned a lesson that would stick decades later.

This is why AI is not going to take over the world. The situation proposed has to be one of the most preposterous I’ve ever seen. NO stakeholder values innovation over stability. Stability is like gravity. The expectation is that it just “is”. You earn the right to innovate BECAUSE your environment/services are stable. It’s not an either/or choice. It’s a sequence of events. Stable ☑️ = right to innovate

This is the easiest problem to fix that leaders have the hardest time executing. Ready? Don’t shoot the messenger. Encourage people to use the basics like the scientific method to find that elusive root cause. If you want your team to fix problems, or better yet avoid them from occurring in the first place, don’t berate them when they’re trying to fix a problem - give them space and support to resolve the issues. Then when they’ve finished triage and it’s time to dig into the root cause, focus on the problem not the person. If someone makes a mistake they usually know it after they find the solution. Allow your team the space they need so they can make mistakes, communicate them and best of all: learn from them.

This is one of the skills technical people, even technical leaders, fail to master. If you cannot explain the concept you’re discussing in common vernacular, you don’t understand it. People already know you’re smart. That’s how you got the job. Move past it. If you can’t also speak in non-Nerd or at least Nerd Lite you’re going to lose your audience. It’s unfortunate because so many talented and intelligent people have great stories to tell. If you can’t explain it to your parent or grandparent keep working on “finger painting” the message. Break it down. Use analogies and metaphors. Connect with your audience.