|
>I'm tired, HN. I'd like to make at-least-ok software in 40h/week and have a life outside of my work laptop. I didn't go that route because I'm probably not in your situation. We build custom data products for enterprise. I wanted the CEO to find clients, close deals, and network. I wanted the CTO to make technical decisions and pick from a collection of implementations. I wanted our machine learning practitioners to dive deep in our clients' data and build models. I wanted to leverage their comparative advantage. I wanted people to do what they're good at, be successful, and deliver value for our clients effectively, efficiently, consistently, and repeatably. Whenever I caught our CTO, CEO, or data scientists doing something I could do, I would push them to do something I could not yet do and take care of whatever they were doing. If they were doing something I knew I could stretch a bit and be able to do, I would go for it and ask for corrections, but the bulk would have been done. That meant owning everything as an individual contributor. The code to do something, the docs, the deployment scripts, the issue templates, the design of the product, working on modularity, focusing on "the job to be done", improving development workflows, documenting findings in a knowledge base that became a seed of our knowledge base, working on onboarding. The business side, better communication and marketing strategies, improving ways to help our clients pinpoint the problems, interfacing with domain experts in different functions, proof reading emails and reports, even making LaTeX templates for our reports and invoices because I wanted them to be beautiful. Later documenting how to operate the company in an "Operator's handbook", finding better abstractions. Post mortems and root cause analyses on failures. Why did a project fail, why did another succeed, why did someone quit, are there patterns. Everything is dissected to amortize the experience. We had paid the price, we definitely ought to keep the learnings. [that meant going over email exchanges of past projects and decisions and "replay" them] It required a lot of comparative reading about a lot of subjects and leveraging past learnings from a period I read voraciously. I wanted us to become more efficient and effective in delivering value to those who trusted us. I'm not smart enough to pull that off in 40 hours/week, as that's basically two work days. I worked everywhere, while commuting, during the week-end. I'm not in your situation so that might not be possible, but the core message is to increase the likelihood of people succeeding at their job in a repeatable manner that doesn't need my presence. Increasing the likelihood of success for projects. Helping clients be successful. "Institutionalizing knowledge" for every single part of the company. |