SRE at 53 watching my role get automated – how do I build something of my own?
I'm a Site Reliability Engineer at a large SaaS company. I can see the writing on the wall as we actively train AI to replace what I do. I'm not interested in retraining for another role that faces the same fate in a few years. I'd rather use this as a forcing function to build something independent.
I've been experimenting with AI tools but don't have a clear direction yet. Has anyone successfully made this kind of transition? Looking for honest resources and hard won lessons not highlight reels.
3 comments
Product market fit is hard. You never know ahead of time if your great idea is actually great. (And people will pay for it). I used to laugh at startups that spent a million bucks building something nobody wanted or needed till I tried doing it myself. If you have 3 hours to chop down the tree spend two of them sharpening the ax. (Do everything in your power to validate a product idea before building it)
There’s a huge difference between something you have to convince people of and something that is so obvious everyone thinks they could have thought of it.
If I were you I’d lean into your intuition of how precisely AI will replace what you do. The fact that you’re certain of it means you can picture it. Build it yourself and own it. (A trustworthy Ai SRE is quite a product)
Is this digging your own grave? Maybe, or maybe it’s putting something on the market that people can choose to use or choose not to use. (Some people choose not to use an automatic transmission)
If you build it and you replace yourself you win big. If you build it and it’s still not enough to replace yourself you know where there’s space for you in the coming years.
There’s a huge difference between something you have to convince people of and something that is so obvious everyone thinks they could have thought of it.
If I were you I’d lean into your intuition of how precisely AI will replace what you do. The fact that you’re certain of it means you can picture it. Build it yourself and own it. (A trustworthy Ai SRE is quite a product)
Is this digging your own grave? Maybe, or maybe it’s putting something on the market that people can choose to use or choose not to use. (Some people choose not to use an automatic transmission)
If you build it and you replace yourself you win big. If you build it and it’s still not enough to replace yourself you know where there’s space for you in the coming years.
define build. product? side/night project? consultancy? there will be still a lot of (SRE) work building out hybrid cross-cloud that goes deep into cost and performance optimization - that AI cannot solve (e.g. have direct access to engineering in mid/big tech). if product building is your thing - focus on problems in your domain. make a list. ask others how painful it is vs how much money it can be worth etc. make it fun. my2cts good luck!