If we're talking about "a company that only deploys rarely and requires many manual steps with many people involved to do it", then I've worked with a few. Typ

If we're talking about "a company that only deploys rarely and requires many manual steps with many people involved to do it", then I've worked with a few. Typically they're shops that got very big very fast. They deal with problems of scale where schema changes are very expensive and data migrations can require weeks without heavy optimization. They typically have tried to automate some stuff, but the crush of feature work to support the massively expanding business always came first. They usually have CI, but never enough tests (unit, integration, or functional) to deploy with any real confidence. They used to deploy (by hand) every couple of days because it was easy. Then they just kept getting bigger and it became every week. Then maybe once a month. Typically they'll talk about why they don't do continuous delivery or deployment because deploys are RISKY and the business depends on deploys being perfect. I've seen this situation in video games (twice), a mobile app that had reached the 100 million user threshold, and an old school data marketing company. It definitely is a thing that exists.