That's really part of the problem of doing a bootstrapped or early angel type startup, and I definitely feel this guys pain. You spend years doing very shallow work because you simply do not have the time or money to dig deep into any given issue until it actually crops up, at which point you do the bare minimum to fix it and move on. Repeat this process until you fail to find a viable business model, or find a viable business model at which point if you're really lucky you can start slowly rewriting your app piece by piece, but more realistically you just accept that it was good enough and keep doing more new things.

This skillset just isn't that useful to a large software shop.

I started software development in startups/innovation teams and this is the reason I am telling younger friends who have yet to graduate to go for the big companies.

Startups are fun. I work at one now. 3/4 software engineering roles I have had have been startups. But all the work I am doing is also very shallow because everything is a fail fast timeboxed sort of thing where you work across a large spectrum of stuff and I will eventually need to change that.

One of the non-obvious challenges with doing a startup is that the majority of your work is spent slogging through boring, non-complicated tasks.

You’re simultaneously doing some of the hardest work of your life but also watching some of your skills atrophy simply because you’re spread so thing across so many different problems.

Startup founders tend to make good product managers or cross-discipline engineering managers for this reason, but it can take some time to slide back into the groove of deep technical work in specific domains.

Agreed - anyone applying for a job as a data scientist should at least be able to work with SQL.

That struck me as a red flag too, I wonder if he's really getting passed because his experience is all on stuff thats too small scale. Regardless, it doesn't sound like he actually did any interview prep, so it's not surprising he's getting negative responses. Even people who love this stuff have to brush up on the areas they don't frequently use (for work or fun).

yeah agree. you want to know what kind of query that ORM generated some might be unsafe.

Or debugging. Or performance.

Basically this guy discovered he is not actually very deep but has not actually concluded that from the evidence before him.