News for hackers. Stuff that matters.
Re:Managing Difficult Software Engineers by thrwayaistartup on Thursday August 17, 2023 @08:15AM (#37160423)
Saying that an entire team is unfamiliar with an IC's tech stack is indeed weird in most cases. But it can make sense, and largely depends on the relationship between the team and the person involved.

A good example from a recent gig: one of our scientists had a HUGE pile very complicated MATLAB, based on several dissertations worth of novel work on both numerical analysis and highly domain-specific mathematical modeling.

Our software engineers needed to either call into or directly use the MATLAB code. Sometimes changes were necessary. This caused a ton of friction for two reasons. First, our SWEs didn't know much MATLAB. Second, you'd have to read two or three papers of complicated mathematics to even understand what the code was doing prior to changing it, and most of our software engineers topped out with Calculus or maybe a Linear Algebra course. So our engineers were unfamiliar with both the tech stack and also the "knowledge" stack.

In that case, I think it's more accurate to say that the software engineers were unfamiliar with the scientist's tech stack than the other way around. There's no way in hell them or anyone else was going to come anywhere close to correctly rewriting the MATLAB code in any reasonable amount of time.

You can think of hiring those types of scientists as one-man startups that are bringing in their own tech stack and debt that your existing org has no idea how to integrate. You need to plan that reality into both compensation amount (err on the side of "way over asking", since this is going to suck harder than the scientist things it will and you need them to stick around) and vesting cliff (err on the side of making them stay longer, since this is going to take a while). And you need to bank on them probably peacing out when the cliff hits.

There are no comments on this comment yet.