4 ms·

Any data scientists on HN? I just ask because I wonder if it would be a good career move for me. Spent 10 years as a programmer and systems analyst. Got real

Any data scientists on HN? I just ask because I wonder if it would be a good career move for me. Spent 10 years as a programmer and systems analyst. Got real tired of that.

Rickasaurus 13y ago
It is a good career move, the salaries tend to be about 50% more than the those for great programmers. However, be prepared to spend at least ten years getting to the point where you are as comfortable as you are now. There's a whole lot to learn.
- mililani 13y ago
  What about salaries? It took me forever to scale up to over 100k as a programmer. I'm guessing it's not much different with data science.
noelwelsh 13y ago
"Data scientist" is a really vague term. It's really a matter of looking at your current skill set and interests, and seeing where they fit. For example, many CS grads don't have strong stats/maths backgrounds. However there are still many "data science" jobs open to them. For example, you can work on the tooling that goes into storing and processing data (e.g. Hadoop), or you can build visualisations. If you want to work on, say, data analysis you need to have a stronger background in stats.
- UK-AL 13y ago
  Since machine learning and A.I is pretty much applied stats, there has got to be quite a lot of cs grads with good stats skills.
  - noelwelsh 13y ago
    Some CS departments require undergrads to do the typical engineering maths sequence (Calculus, Linear Algebra, etc.) but many don't. Lots of PhD students in machine learning have undergrad degrees in physics, maths, or engineering. Where I did my PhD (the top ranked CS department in the UK by a recent survey) the undergrad students did one maths courses in their entire degree, with most of it being discrete maths (IIRC). This really limited the ability of students to undertake research in machine learning or theoretical CS.
    - UK-AL 13y ago
      Since your based in Birmingham I can take a good guess at where that is, because I studied there and live in birmingham. And I agree, but the machine learning/A.I modules themselves had to have maths placed within them. Nice to see brummies on hacker news. Since your start up guy, I guess you hang around faraday wharf?
      - noelwelsh 13y ago
        I was at Faraday Wharf a few months ago, but I mostly work from home now. Drop me an email (address is in my profile) if you want to chat more.
altrego99 13y ago
I work for a big firm doing data science - or risk, fraud, response etc. prediction specifically. It's easy money. The problem that you can run into is this. The systems are archaic. By that I mean both the implementation platform, and the people who you will sell your solution to. To most of the senior stuff this is new and mysterious - they would rather "play safe", and may ask you instead to use traditional methods, put "intuitive" variables, and keep the solution "simple" (simple being anything that they already know of, from their past 10-20 years of experience). But things are changing. In my farm too. And if you interview now as a data scientist, you will probably be given the right respect - and put into a team where the leaders may want people like you. And you will be one of very few in the company who knows what this stuff is, rather than myriads who know how to model using the company's already setup platform - which you will prove to be inferior. You may have to push the boundaries of the company for it, and you will pave the way. But you will also be welcome - as the realization now dawns upon corporates that there may be better ways of doing what they have been doing. Also you will not be alone. You will find some smart people in the company you join soon. They will challenge you, be your friends. You will learn and think about things that you have not before... as will you intrigue others to think of things that they haven't before. So go for it. Try it out. It's easy money.
- mililani 13y ago
  I hope you find this. I was wondering if it's ok to ask you what "easy money" means? Quantitatively, that is. Care to share ball park figures? I can never get accurate or reliable figures with salary sites online.

Any data scientists on HN? I just ask because I wonder if it would be a good career move for me. Spent 10 years as a programmer and systems analyst. Got real tired of that.

Rickasaurus 13y ago

It is a good career move, the salaries tend to be about 50% more than the those for great programmers. However, be prepared to spend at least ten years getting to the point where you are as comfortable as you are now. There's a whole lot to learn.

mililani 13y ago
What about salaries? It took me forever to scale up to over 100k as a programmer. I'm guessing it's not much different with data science.

noelwelsh 13y ago

"Data scientist" is a really vague term. It's really a matter of looking at your current skill set and interests, and seeing where they fit. For example, many CS grads don't have strong stats/maths backgrounds. However there are still many "data science" jobs open to them. For example, you can work on the tooling that goes into storing and processing data (e.g. Hadoop), or you can build visualisations. If you want to work on, say, data analysis you need to have a stronger background in stats.

UK-AL 13y ago
Since machine learning and A.I is pretty much applied stats, there has got to be quite a lot of cs grads with good stats skills.
- noelwelsh 13y ago
  Some CS departments require undergrads to do the typical engineering maths sequence (Calculus, Linear Algebra, etc.) but many don't. Lots of PhD students in machine learning have undergrad degrees in physics, maths, or engineering. Where I did my PhD (the top ranked CS department in the UK by a recent survey) the undergrad students did one maths courses in their entire degree, with most of it being discrete maths (IIRC). This really limited the ability of students to undertake research in machine learning or theoretical CS.
  - UK-AL 13y ago
    Since your based in Birmingham I can take a good guess at where that is, because I studied there and live in birmingham. And I agree, but the machine learning/A.I modules themselves had to have maths placed within them. Nice to see brummies on hacker news. Since your start up guy, I guess you hang around faraday wharf?
    - noelwelsh 13y ago
      I was at Faraday Wharf a few months ago, but I mostly work from home now. Drop me an email (address is in my profile) if you want to chat more.

altrego99 13y ago

I work for a big firm doing data science - or risk, fraud, response etc. prediction specifically. It's easy money. The problem that you can run into is this. The systems are archaic. By that I mean both the implementation platform, and the people who you will sell your solution to. To most of the senior stuff this is new and mysterious - they would rather "play safe", and may ask you instead to use traditional methods, put "intuitive" variables, and keep the solution "simple" (simple being anything that they already know of, from their past 10-20 years of experience). But things are changing. In my farm too. And if you interview now as a data scientist, you will probably be given the right respect - and put into a team where the leaders may want people like you. And you will be one of very few in the company who knows what this stuff is, rather than myriads who know how to model using the company's already setup platform - which you will prove to be inferior. You may have to push the boundaries of the company for it, and you will pave the way. But you will also be welcome - as the realization now dawns upon corporates that there may be better ways of doing what they have been doing. Also you will not be alone. You will find some smart people in the company you join soon. They will challenge you, be your friends. You will learn and think about things that you have not before... as will you intrigue others to think of things that they haven't before. So go for it. Try it out. It's easy money.

mililani 13y ago
I hope you find this. I was wondering if it's ok to ask you what "easy money" means? Quantitatively, that is. Care to share ball park figures? I can never get accurate or reliable figures with salary sites online.