People sometimes don't think about that. They think about job titles and they're like, whoa, you know, if we want to do cross-functional backgrounds, you know, we need to get a software engineer and a data engineer and some data scientists. And sometimes I've told people, I'm like, you need to get people from different backgrounds too.
You can't get a whole bunch of people that grew up in Manhattan, in New York City or a bunch of people that grew up in, in Palo Alto and expect that they're going to have differing viewpoints. They grew up in the same environment. They matriculated through university at probably the same school. They didn't get exposed to different ways of thinking through things. Get people from different backgrounds, from different countries, from different cultural backgrounds.
They will make a much better product. And that's been proven out, I don't know how many times, in Silicon Valley of looking for that diverse global talent pool and picking people from different backgrounds to put them together and they make magic happen. Same thing should be done in a data science team as well, because I would argue that it's even more important for data science use cases because software engineering is creative in the sense of you're solving logical problems in unique ways, but there's a lot of disciplining and structure around that because you have to. It's got to work. It either works or it doesn't work and it's got to work. So you need a lot of controls around that. Data science and even some aspects of ML engineering: there's a lot of creativity in there. Sometimes you're going down a path that nobody's tread before. You have this unique idea of how to do this thing with data or how to combine this feature engineering processing with this model architecture. So it's a very creative process and it's also a business focused process.
A software engineer is building something for use by a client or internally at their company. And they're almost like a consultant that works within the company. You're just delivering something that people are asking for and it has to meet these product requirements and it just has to continue to work.
Whereas data science is, Hey, we have this problem we need to solve. Nobody knows how to solve it. If they did, they wouldn't be talking to you. And it's very creative and potentially frustrating because you're going to try a hundred things and ninety-nine of them are going to be wrong. Hopefully one of them is going to be correct and you have to creatively do that.
So the more diversity that you have in technical background, as well as cultural background, you just get different viewpoints on how to solve something. And I've always found those teams solve more interesting things faster and more reliably than a team of monoculture to mono-job-title, mono-background individuals.