Do you hate your data science job?
Data engineer, data scientist, data analyst, data researcher, perhaps even ML engineer, whatever title you have. Also tell us why yes/no if you have a minute!
Hive queries are so fucking slow. Hive queries are so fucking hard to test and debug. Hive and Presto SQL have so many fucking shitty annoying differences. Google, save us! Such fucking shitty Hadoop / Hive. Fuck Yahoo!!!! Fuck Apache!!! None of the fucking Apache projects are super user friendly. They all seem fancy, but quite unpolished.
No because I'm very good at what I do, training/mentoring other data engineers and scientists in addition to speaking at conferences/events.
Truth
Yes, I hate working consulting hours. Business majors need to realize hours of work != impact, productivity
In my team leaders don’t understand ML and give me stupid projects to work on. Like prediction attrition based on address of employee. It doesn’t meet common sense bar
Actually like it despite being at the worst of the worst bodyshops. Meaty DS project and it's not way over my head
Hate would be too aggressive a work to use. I would say I don't like my data science job, for the following reasons: 1. There is no clear growth picture. Since it is a new field, the leadership at most companies doesn't know what is the hierarchy and growth responsibilities of data scientists. 2. Most managers who did not grow in an analytics/data science role and transitioned from other roles say software engineering or IT, they think building a regression model is just about writing an lm function. They do not understand the data cleaning part is 60% more time consuming than writing a model. Similarly, improving a model is a whole lot effort in itself. 3. You not only have to be good at data science and ml algos, but also have great communication skills. Because not everyone is technical enough to understand your models, and you need to dumb them down to an extent that it loses its value 4. The expectations of getting into an actual data science role at great companies (not the BA roles sugar-coated as DS) is pretty high. You have to be good at algorithms, programming, business sense, product sense, communication, and a variety of tools. 5. Most of the time is spent in data gathering, data cleaning, and presenting insights. The core data modeling work which is the most interesting part (to me) is only 20% of the job
20% of model building is the highest (and most generous) I've heard yet. Data (software) engineering is the bulk of the actual work, usually around 90-95%, so kudos for being able to spend even 20% of your time there. I would say that knowing your data, which comes through engineering and analysis can be the most important aspect of your job IMHO
Wow, thanks for taking time to let me (and everyone) hear your thought! I agree with you. #4 is what motivates me to create this poll - tech consultants throw big words like big data, data science and AI out there and bring little value to the actual table....
I love it. I am on the application side of it. It's about trying new ideas, understanding domain, building features, trying ml, dl. Though I have not contributed to the research but have been successfully able to deliver the business value by using those research. The best part is I get to think over the problems a lot. The combination of programming, stats, ml/ai, maths, big data make it diverse and quick. Delegating thinking ability to the machines is the key too. Some examples when word2vec was released for nlp, I applied it in click recommendation..sequence of clicks on product ids are like sentences :) Anomaly detection using lstm :) I feel motivated when after solving a problem I can see some research already going on exactly using the same ideas. With databricks, spark, python, keras it's beautiful :)
Lol all these yeses, tell us why!