I have been working as a data scientist now for about 3 years and in this video I want to shed some light and explain the brutal truth behind what it's actually like being a data scientist let's get into it so what actually is a data scientist well nowadays it can mean a range of things but I like to think of it as a data scientist is someone who uses programming maths statistics and data to help solve business problems and provide insight into where things can be improved or how to use data more efficiently to me
the main unique skill of data scientist is that of building machine learning models and predictive statistical models we do these to help Drive business decision- making and make process a lot more efficient than they would if they were just operated by humans common problems that data scientists may work on is predicting if a credit card payment was fraudulent or not recommending the right products to certain customers on e-commerce websites and also generating the right kind of insurance policy for a customer data science is such a big area so what you find is that as data
scientists get later on in their career they will often specialize in one particular area like recommendation systems reinforcement learning and forecasting and they will also probably specialize in one particular industry domain so the work I do as a data scientist is that I work in a cross functional team and my team mainly focuses on optimization and forecasting type problems much of my day-to-day work revolves around trying to improve the models and also providing support to stakeholders to explain why the models are doing a certain thing the general workflow is that someone will have an idea
about how they can improve the model will then go into trying to find if the data is readily available to prove this you know idea or hypothesis if the data is available then we'll go into a research phase so we'll build this idea in the model and we will test it then after we've the test we'll then do some analysis basically just to determine has this idea or hypothesis actually improve the model or not then if it has we'll then go to ship it into production by writing production code and making it nice and tidy
and then sending it out into the world now along this process there's a lot of collaboration with other teams and functions for example when we're doing the initial idea and scoping of the hypothesis this may require us interacting with product managers to get that business View and to make sure we haven't missed anything on that business critical side getting the data research this can all be done in tandem with data analysts and data Engineers because they have access and better knowledge normally of the data and the pipelines just make sure that we are using indeed
the right data and we're building the correct ETL pipelines and finally the ship stage is often done with software Engineers to make sure the way we're deploying our algorithm is indeed up to kind of production code standards and is the best way of deploying an algorithm from my general experience and chatter with other data scientists this type of workflow is pretty consistent across the industry so if you are looking to become a data scientist this is kind of the general way you would expect to work in a company data science teams are structured in kind
of two main ways incorporations and the other ways are kind of just like a mixture of the two the first one is the idea of having embedded data scientists this this is kind of how I work in my cross functional team where basically you have a team full of data scientists analysts product managers software engineers and in this team their goal is to basically look after one area or serve a single purpose the other way around is kind of the complete opposite and that's having data scientists as almost an inhal consultancy and what I mean
by that is that you'll have a team of data scientists who are kind of like in their own team and then they will be kind of set on tasks inside a company which uh become the most value at that point in time this kind of system means that you work on a range of problems so you won't necessarily be a specialist in one domain you may be uh assigned a forecasting problem one day and then a reinforcement learning the next day so both these kind of systems have their pros and cons but to be honest
it doesn't really matter because I find the work so interesting no matter what kind of structure I have I enjoy it nonetheless I found find that people often really glamorize working in Tech online like it seems to be all coffee chat coding an hour a day and yet somehow you make six figures in reality this is not the case but the work day is still very interesting compared to a lot of other professions my general experience is that the workday typically starts at 9:00 where you have a morning standup and then 9:30 till 10:30 you'll
have some sort of sync or meeting with you know product managers stakeholders they to scientists just some meeting to discuss recent developments then between 10:30 and 12 to 12:30 you'll have a work block where you can do some coding then 12:30 to 1:30 you'll typically have your lunch break and then after that you have another coding block from 130 to about 4 and then 4 to 5: you may have another meeting or sync with stakeholders then from 5 to 5:30 you basically just wrap up the day answer emails slack messages and kind of get all
organized for tomorrow it's important to mention that we don't only just code as data scientists we're kind of like the Lynch pin between the business and Tech side so some days I may be doing presentations for stakeholders and explaining what my data and my models are really doing to the business the reason I'm a data scientist can be boiled down into four main reasons the first one is that the work is just so interesting I mean we're currently in this big AI Revolution so as a data a scientist I'm pretty much right the Forefront of
all the latest developments in machine learning AI deep learning whatever you name all these buzzwords and it's just really interesting because data signs like I said earlier it's such a big field so you can always bet there's something new to learn and you can never learn everything and that idea of constantly developing learning new interesting things to improve your career it's something that I really love the second one is that the flexibility and the work life balance tech jobs are generally quite well known to have better work life balance than other careers like banking or
law most kind of jobs I've seen for data science work no longer than 8 to six but most of them are normally around 9 to 530 and with that a lot of tech-based jobs normally have a quite hybrid working policy so you can expect to work kind of a few days or a couple of days from home in the week I personally love this because all that time I would have wasted commuting or working till 10:00 if I was a lawyer Banker I can spend pursuing my other passions such as creating YouTube videos writing blogs
playing hockey just gives me a lot more variability and kind of freedom to pursue other Ventures that I'm also interested in as well as data science the third one is that the compensation in data science is pretty good I mean it's pretty well known that Tech generally plays quite well particularly for North America where it's not unheard of for data scientists to be earning 200k Plus in the UK in the EU it's slightly less but it's still above average compared to the the general median salary of the region and the final one is that as
a data scientist you can work in so many different Industries I've only worked for 3 years and I've already worked in both insurance and e-commerce so if there's an industry that you're really passionate about then chances are as a data scientist you can probably get into it fairly easily so the final point is that should you be a data scientist well of course you should but only if you want to data science is a great career like I've just said but it's not easy all the time and if you don't like coding maths statistics and
constantly learning things then it may not be the right fit for you I have a whole separate video explaining exactly the things that you should be comfortable with if you want to become a data scientist you can find it linked on screen here if you enjoyed this video make sure you click the like And subscribe button and I'll see in the next one
