 Hello everyone and welcome to another episode of Code Emporium where we are going to talk about the data science career ladder. Now it's kind of a career ladder and more of a career trajectory because data science itself is a very loose field. We have people of different backgrounds coming into the field at different levels and we also have people from different levels in data science also going to other fields. I thought that this would be a good video to kind of give some sort of structure to the entire process as a whole so that we could kind of see at least some sort of trajectory or direction of where you would potentially be heading if you were to take up a data science position today. But before we talk about that please do give this video a good old like. Also please do check out the discord server down in the description below. We're talking about some amazing things and we would love to have you as a part of our wonderful growing community. Ring that bell and with that let's get back to the video. So I've created this little schematic representation of what I believe a data science trajectory should look like at least adding some structure to it. And on the left side here I've also introduced a gradient increasing arrow where it says broader problems on the side right here. And we have smaller problems that are at the task level that are kind of dealt at the entry level data scientists. And then they kind of expand to more holistic project based levels than multiple project based levels. And then finally when you hit an executive level you hit organizational targets and basically trying to see what direction we would take an organization in. I thought this would be just a good other perspective of how to look at what kind of tasks data scientists do at different levels. Because in the end data science is about solving problems. And if we think about abstracting the role of data scientists at different levels as different scopes of problems then I think it also becomes a little easier to understand. So starting off we have the entry level data scientists. Now these are data scientists who could either be interns or they are fresh graduates ready to start their data science career at a company. They have about zero to one year of experience in data science. And the type of work that these data scientists do are associated with typically a particular task. So maybe they'll be asked to go into a repository of an existing data science service and are asked to modify a given feature or maybe add a small feature as a part of that service as a whole. The entire idea here is just to make sure that as an entry level you become aware of the data science tooling that is available to you either from model development from data analysis to deployment and also understanding some of the kinks in some of these like preprocessing techniques and kind of basically transferring your knowledge of whatever you learned in school at least a little bit to what you can use in the industry. Once you kind of get the knack of different stages in that data science pipeline you can get to the mid-level roles. And this is the current role that I am in right now as a data scientist with two years of experience. And you can expect to be around like one to three years of experience at this level. The kind of trajectory that I have seen from an entry level to a mid-level is that now I am asking a lot more questions. I am also trying to tackle projects end to end from start to finish. And it's at this stage where I see machine learning less as just a statistical method but more as just a mode to solve a problem. And setting up machine learning pipelines becomes pretty perfunctory to me. I can kind of do it off the fly without much of a hassle. At a mid-level data scientist role at least for me I have this mindset of solving problems and so it makes sense for me to come up with a problem come up with solutions to that problem and then also ask the advice of some of my more senior data scientists just to see what they think if I've missed anything at all. Now after spending a couple years probably in the mid-level role you might see yourself in two different paths in the future. One is an individual contributor path and the other is a management path. So by individual contributor you are mostly focusing more and more of the same of solving problems but at a larger scale as this arrow indicates. And in the managerial path you'll mostly be focused on not only solving problems but a lot of your work is also trying to be responsible for the growth of mid-level and entry-level data scientists as well. They make mistakes, they have a lot of questions, they need advice and they go to you as the data science manager for everything. Now typically in certain companies you would want to actually become a senior level data scientist before you see this branching but I've just put it over here directly after mid-level just to avoid confusion. After all a lot of these roles kind of encompass multiple roles too and there's no fixed framework for every single company of how they decide to promote candidates and this is just a general frame of reference to keep in mind. So now coming to the role of senior data scientists slash manager which are kind of in similar positions to each other in terms of the scope of problems they solve. So as a senior level data scientist you have a few years of experience under your belt, you are able to solve problems taken from start from problem ideation all the way up to the deployment of your problem and you can do that pretty seamlessly without much dependence on anybody else. It's also at this level where you start seeking your own problems trying to see where there is opportunity, try to come up with opportunity estimates, get buy-in from maybe dependent stakeholders who are dependent on that. By buy-in I mean you come up with like a little presentation telling a couple of stakeholders hey this is where you can save money and I also know a way of how you can save money so you want me to help you save money and then the stakeholders will give a yay or an a to you and depending on that if they say yes please do save money for us then you can go ahead and implement your pipeline start to finish communicating with stakeholders which is key entirely but without much dependence on them you probably might just be asking them for certain questions just to make sure that the system works in an accordance to how they want the system to work but all in all senior data scientists super independent and super reliable. Now I put the manager role here too because a manager does also go into the details a little bit but they are also like I mentioned before responsible for the growth of mid-level and entry-level data scientists which do take a chunk of their time. A level above this would be the principal data scientists as well as directors. Now you can see for these managerial roles and the individual contributor roles I put a little arrow and this basically means that you can kind of have a good balance between each of these depending on the company that you're in and also depending on the nature of data scientists that you want to be. Some more individual contributors are not just individual contributors they would also act as mentors to people who are mid-level senior level or entry level data scientists and then those who are managers aren't just managers but they also wanted to get into the details of solving problems end to end. So when talking about a principal level data scientist I honestly do not have much intel on this because I do not interact with them on the daily however we can look to the internet for some solutions on what the work of a principal data scientist entails and also that of a director what that entails. So here's a question on Quora how is the work of a principal data scientist different from that of a senior data scientist in large tech companies like Microsoft and Amazon and we have Kristen who is a principal data scientist who answered this question basically from this nicely written answer here they basically say that a senior level data scientist is able to execute tasks end to end but they typically do this one task after another in a sequence whereas at the principal level they would be working on multiple tasks at the same time some projects might be in the development place other projects may be in the deployment phase and at the same time they might also be mentoring some senior level data scientists entry level or mid-level data scientists as well and given a problem these are the people who are most likely to be able to solve it if there's nobody else in your company who can tackle it in a very proficient way now coming on to the director position we have another core answer who's answered this question on what is a better title chief data scientist or director of data sciences now Sean McClure is a director of data science here and he's also given a pretty wonderful answer to this he starts off with saying that a chief data scientist who is kind of like a principal data scientist is involved with the same day to day tasks as the junior members of their team so they're involved with a lot of individual contributor work but the director of data science tries to direct and execute the most important work and strikes the right balance between talents and helps coach team members on how to tackle problems at scale and align efforts to the bigger picture so they're kind of like the conductor that kind of sews all the team together this is really interesting because actually as a director and also a manager you will really be in charge of just dictating the kinds of problems that everybody solves so you need to strike the right balance between hey we want to solve problems that are very useful to the company but we also want to make sure that the problems that we are solving are also interesting problems from the perspective of a data scientist striking that balance is tough because there are many useful problems that are extremely boring and there's also many many data science fun problems that are useless to a company depending on that industry that they're in and it's here that the manager either at a lead level slash managerial level or the director level is going to help strike that balance and tries to push forward on certain problems that we solve now he's also given a pretty cool analogy in a hockey team where the chief data scientist the principal data scientist would be kind of like the captain and then the director of data science is more like the coach using their knowledge and experience in playing the game to see the bigger picture and find the chemistry between the right players to consistently score the most points i really do love that analogy and to close things off he says that the lines between them can be blurred so sometimes he would be in more icy work rather than just managerial work or sometimes it's flipped where they are looking over senior level data scientist entry level and mid-level while also trying to get their hands a little tinkered into the to the weeds whenever they feel like they want to as an additional level i included the executives up here um executives is definitely not just a position that oversees data science but it could be an it could be a position that oversees an entire organization this involves some of the c-suite roles kind of like the chief technology officer or it could be a vp level but in general um i'm not really going to talk too much here because their roles could be managing not only data science but it could be all of like data analytics it could be engineering altogether too but this is a good example of where it is possible to go from data science to be in a much more broad role that touches various aspects of an organization also to close things off here i put some gray lines coming from these clouds with other in the middle of them this basically is to show that data science itself is a very welcoming field for people of various backgrounds you could have a software engineer background or a sales background a marketing background and just be data savvy and use the problems and experience in solving problems that you had in your past experience to actually fuel your knowledge as a data scientist because those problems can be solved also from just a different perspective of using data and hence a lot of those skills are transferable and so i say okay from other categories you can become a mid-level data scientist a manager or even like go to the executive level too and that's kind of all i have for our data science ladder or more a data science career trajectory it is to show a trajectory of how you would start from maybe start to finish but in many cases if you look on linkedin you'll see that a lot of people who are data scientists they were something else in the past and they might become something else in the future that is outside the scope of what i've written here but it is definitely always a fun space to be in when you're there at the moment and so i hope this video does help you just understand the overall structure even a little bit it really does vary depending on the company the size of the company the type of the company too but yeah i hope this helps please do drop a like down below also subscribe for more amazing videos from yours truly and i will see you in the next one join that discord now i'll see you in the next one bye