Data Science is more than just Data Scientist - Different Roles in the field of Data Science

Key Takeaways

hi everyone in this video we will talk about the different kinds of rules that you can apply for in the field of data science this will give you an idea of how broad the field is and perhaps help you find your dream job different roles you can apply in the data science industry itself so the data science industry now has multiple roles in if you what I can what I like to call data science industry is all the rules that are involved being the use of data so these are the kind of roles that you would typically have some or the other involvement of data as the first one is the business analyst role where you're essentially trying to understand the business objectives you're trying to see how you can contribute to business and then conveying those problems into maybe a greater you may be kind converting that business problem into a data problem and conveying it you're more technical teams so you are kind of looking out on how what are the user is doing how many users are there how can i how can i account for churn if they're joining how can i look for more growth of avenues using data right so this is what you would do as a business analyst at company then there's this data analyst role that also there's also something you can apply for and this is also involves you to work with data and this typically requires you to know some sequel where you are writing sequel queries getting data out of databases and then maybe using that data to run descriptive statistics on top of it maybe you are presenting in the results in the form of a report or a dashboard to say for a business analyst as well you might also have to do reporting dashboards that but data analysts a little bit more technical so then they are able to write sequel queries some programming sometimes some Python or are but typically your row is to do a lot of descriptive statistics as a data analyst then comes the role of data engineer who takes care of the pipeline's that of data in your company right so essentially data exists in a lot of different places I mean I'm talking about as their company grows if you're if you're part of a bigger company then the data tends to grow in silos because there are a lot of different departments and these departments don't talk to each other often and you have valuable data stored in different departments you in different kinds of systems as well so how do you basically integrate all of this build a pipeline and maybe put it in a data lake or a rate of warehousing system and allow the user your your end-users here would be typically a data analyst or a data scientist but allow them to consume that data right so as a data engineer you're responsible to build those pipelines and in fact even a lot of these roles are big data engineering roles right like Hadoop or Kafka or all of these things these technologies are used by data engineers because the moment you start having tons of data coming in you need systems that could reliably process them and store and data engineers work on that so yeah data engineering role is also in fact a very widely available role in the industry not a lot of people realize and this is one role that does not require a lot of mathematical skills because this is more of an engineering role where you're building pipelines where you're seeing what kind of databases to use what makes sense what doesn't make sense so data engineering is something that you can still look at if you are if you're not keen on the math side of things if you're not keen on yeah if you're not keen on actually doing the extra work credits will provide for that so the data engineer is a good role that way and these are also generally well paid roles so that is something you can also look at then you have the machine learning engineer role who are typically looking at improving the performance of their models and trying to tune their the hyper parameters of the models are trying to get the best accuracy or best f1 score or best ROC curve or whatever metric they are tracking right so machine learning engineers typically are looking to build systems that involve the use of machine learning models and their day-to-day job is improving the models that they're using maybe by improve improving the way that data is represented or by tuning that like I said have a parameters or by choosing a different models of this specialize in that area of the field where they are they're responsible for improving the performance of the model day in and day out research engineer research engineers are those kind of these are far fewer role these kind of roles are far fewer because research engineers typically work in researchers typically work in industries in universities and research engineers are typically found in bigger companies like Google or Facebook or your Microsoft because they can afford to do research at that scale so research engineers typically try and see how they can improve the current state of the art and build better models or build new techniques to do existing tasks right like for example if you're doing machine translation you might be using a statistical machine translation model or you might be using a neural machine translation model or let's say you are using an LSTA model how can you build something else right which is not even a part of these families right where you are introducing new research and performing existing tasks better so you're trying to push the state-of-the-art for the current techniques this is one example but a lot of research engineers working on different projects so yeah like for example I think Ian Goodfellow who he works on ganz a lot generative original at work he works at Google right so he is an example of a research engineer he doesn't really work on a product he works at how he can improve Gans itself then we have data visualization engineer or bi engineer who typically work with again a lot of sequel or again one programming language to get the data out and visualize it using one of the BI tools like bi or tableau or QlikView or or all of these tools right so they are also very much involved with data in their day to day activities so they are put together with a lot of other data rules here and finally data scientist which is essentially a combination of 1 plus 2 Plus 3 plus 4 plus 5 plus 6 because you have to understand the business requirements you have to be able to understand you have to be able to understand the business requirements translate that you sequel or any programming language to analyze might have to build some data pipelines yourself build the model yourself as a machine learning engineer would do look into the latest research and try and get some research implemented so that also could be a part of your role as a scientist and then present your findings in the form of a report or a dashboard so data scientist basically is a combination of all of the above roles at a lesser price point if I can say that but yeah so data scientist is since all of these put together but again you are not specializing in one of them you are good enough in a variety of these different categories if you liked the video please do give it a like comment something interesting or share thank you