To survive as an AI engineer, you have to own the solution end to end. One end is the need and the other end is the outcome in production. And you must know every step in between. There is no excuse for not understanding how a piece works. That is your agency
@techpoto@owengretzinger Custom built! Orchestrator service listens to GHA. Delegates to VMs provisioned with custom setup in Azure. (Orchestrator deployed azure contrainer app).
@_anonpenguin_@owengretzinger I understand the appeal of non-cloud solution, but all are infra is cloud hosted and it has other benefits that outweighs the more cheaper alternatives. It is always the trade of where we spend the time.
Not sure it is just "caring". When an engineer thinks they can do better in an area and creates a PoC, they obviously care. What they might miss is:
1- Do they have the full context?
2- Can they spend their "care" elsewhere (while another eng is working on the problem area). The examples in the post (which is not AI generated) is about focusing on solving the right problem.
And these are hard decisions. Maybe the PoC is a great area to focus and improve.
I recently shared a thought that in a world where building software is practically free, the mistake is not to build. I still strongly believe that. But just because we can build almost anything now doesn't mean we're working on the right things. https://t.co/yHkaV44CaP
The @runneth MCP is now LIVE for everyone ⚡️
This MCP is the first of it's kind. Many people told me this approach won't work but they are wrong.
We were early betting on agents having their own computers.
Now we are betting on agents working closely to accomplish ambitious goals together.
Out of the box @runneth is exceptional at marketing.
So having runneth on your team of agents has the same effect of having a star employee: it levels up those around them.
The runneth MCP has one main tool call: ask_runneth.
It's the exact same runneth that you work with in Slack, and now it can work directly with your agents too.
Our team and early users have been having a BLAST using this.
Send your agent to https://t.co/sUtvNp9OnK to learn more.
Every day I wake up to 10-15 scheduled runs on systems I monitor. The next two hours of reading and acting on these reports are my most productive hours. I always had dashboards and reports for the systems I'm responsible for. With AI, it's the next level.