Information Science & Software program Engineering – sustaining each?
I’ve learn a ton of various articles, opinions, and concepts on selecting between a profession in information science or a profession in software program engineering. The consensus on this matter normally boils all the way down to “choose the one that you just take pleasure in probably the most”, which I feel is terrific recommendation.
Nonetheless, I typically discover myself torn between the 2. I am going to spend all of my time constructing information pipelines, modifying information processes, and doing desk joins whereas longing to write down a single line of code. However then I will be adjusting the structure of a cell app or webpage pixel-by-pixel wishing I might analyze an issue utilizing real-world information and ship a considerate evaluation and proposal on my findings.
I am at the moment working as a senior degree information scientist, though my duties replicate the function of a knowledge engineer. I routinely apply software program engineering coding issues on Leetcode and construct apps in my spare time. When contemplating making use of for a brand new job, I might be open to both DS or SWE positions, offered the DS function would make the most of machine studying and constructing precise fashions. On the finish of the day, I simply need to construct one thing that has worth and requires me to consider attention-grabbing options and implementations.
Has anybody else skilled this push-and-pull between DS and SWE? Are there alternatives in DS which have extra of an emphasis on ML and coding that I am simply not experiencing in my present function?
Comments ( 3 )
I’m not really sure but have you looked into google/meta/other tech company’s research divisions? Basically figuring out how to apply stuff like machine learning algorithms in their business.
I did software for years, then DS as it developed in my field. Having done both now, I would choose SWE. I am a consultant though so most of my work is dealing with people who don’t know what anything does, hires a team they could get cheap but have no synergies, and they want nothing to do with listening to outside people because they took some classes and think they know better. To that end I do a lot of R&D but the rest is cleaning up the messes of billion dollar companies who hired poorly and couldn’t complete the job or more that it looks like they got all of their experience from SO.
I enjoyed the challenges of SWE much more. I think my next role will be something combining all of it for a small company, I like components of each. If I can never have someone clueless about DS and stats ask me about hyperparameter tuning in our first meeting that would be great.
A legit 30 billion company came to us to fix their issue, they were doing PCA and thought they could just remove anything that wasn’t in the first couple of components. The guy in charge of the team only knows business, all of the people in his team on the DS side were less than a year out of school and didn’t know how much of what they do works. We fixed it and that guy got a promotion.
SWE or both.