Hacker News Reader

Comment by opportune | Hacker News Reader

That’s why you become a data engineer and get the best of both worlds. If you can market yourself to a company as a “full stack data developer” you will make more money than either category

View on HN

"data engineer" is often considered a lower prestige title than data scientist. "Data scientist" is a title recently thrown around a lot for positions that used to be called "data analyst", with no strong ML or SWE ability required. Amazon's title for scientists with strong ML engineering ability is called "Applied Scientist" and it's paid significantly higher than SWE

You're claiming to know amazon internal except you're also using SWE, not SDE, the standard name for software devs there. Also amazon seems to hire very few Applied Scientists and thus is likely cherry picking for this role. The one I work near is an industry thought leader at Principal Applied Scientist. They'd be a leading researcher if they were in academia.

Software engineers are given the SDE title at Amazon but I just used SWE as the more general "software engineer" acronym.

a title recently thrown around a lot for positions that used to be called "data analyst"

Agreed. Maybe this is different in different markets but in London the vast majority of "data scientist" positions advertised are really PowerBI, Excel, etc. If there is any ML it is just to feed data into a black box model. If you were both smart and lucky you might be able to sneak R in by the backdoor and start doing actual data science, but it would be an uphill battle.

Data Engineer role is about building big data/streaming pipelines for data and often DevOps aspects of keeping such pipeline deployed and online. A boring, dead-end job. Avoid.

Make yourself both ML/DL master and SWEng and you'll be doing extremely well.

I do not find my Data Engineer job either boring or dead end, but thanks for the condescension.

I love data engineering - atleast when I practiced it at an adtech company, it involved lots of interesting challenges around performance, distributed systems and a mix of SWE & Ops work. Buut, applying to data engineering jobs I see where the parent poster is coming from (minus the condescension) - a lot of them are about writing simple ETL jobs, and they tend to be lower paid & less prestigious.

Those certainly exist, but aren’t what I was referring to in my post. The “data engineers” I’ve worked with (and am) are usually distributed systems people who also know a bit of data science. In that case it’s basically a higher “prestige” software engineering role

You know that's usually the hard part of ML projects, right? It may be boring, but it also tends to be the less least replaceable component.

Yes, and there is zero prestige within company, awful lot of high-pressure work, no control over bugs in infrastructure pieces that tend to blow up unexpectedly in production, getting stuck on a certain stack that tends to age badly (who remembers Hadoop or Spark 1.3 these days?), making one quickly replaceable by younger folks with the latest hyped tech, small understanding of the data science part, no joy from constructing great algorithms either. It's a needed job of course; it might pay well at this very moment but it's a perfect target for automation once latest cognitive automation research bubbles down to production.

If you want to shine, do Deep Learning (research if possible), low-level distributed systems you have control of as an author of an important piece of infrastructure or the actual data modeling and predictions.

He's right about how data pipeline work is regarded by most ML scientists. It often ends up being extremely hard work and is regarded by others as useless grunt work. There is no big reward, constant overtime due to unexpected issues and lots of stress.

SQL skills are expected of every data scientist and backend engineer, but are also found in a much larger and cheaper labor pool. As far as I can tell, the data engineering role exists to take advantage of those economics.

ETL pipeline queries and configs are absolutely critical, make or break an ML project, and account for most of its labor hours. But they are the most commoditized part of it.

Reply on news.ycombinator.com