What Professional Data Scientists ACTUALLY Do

We often see the best parts of being a data scientist in day of the life videos. In this one I tell you what the whole experience is like. I want to talk to you about what data scientists ACTUALLY do, not about just the fun parts of the job.

1) There is no one size fits all data science job. You can find what you want out of the variety of opportunities in the domain
2) Project management for data science is a real thing, you're going to have to get familiar with the tools of the trade to have a good experience
3) Different types of projects have different expectations. You have dramatically different responsibilities if you're working on long term or ad hoc data science projects
4) Git and versioning is a must learn for working with data science teams
5) Learning to work with production and development servers is a big difference from personal projects
6) Tools and languages you use vary dramatically by team
7) Work is not always as exciting as you expect, your main responsibility is to drive value to the business
8) Ambiguity is a major part of the role
9) Not everyone loves data scientists in your organization
10) You sit down a lot
11) You get to work on mostly interesting projects
12) You are surrounded by learning and opportunities to expand your knowledge
13) Work / Life balance is usually pretty good

#DataScience #KenJee
0:00 Intro
0:50 Variation in Data Science Opportunities
2:23 Project Management for Data Science?
3:42 Types of Data Science Projects
5:38 Do you Collaborate?
6:33 How SharpestMinds Can Help
7:25 Quality Control
8:00 Can you choose what tools you use?
8:41 Types of Data Tasks
9:40 Be Wary of These
11:27 The Good Parts
13:05 Sharing Your Awesome Projects!

Partners & Affiliates

MORE DATA SCIENCE CONTENT HERE:

Check These Videos Out Next!

My Playlists

[georgeventura]

Im 35 years old, and I just started my Data Science journey by starting college this semester. Been an aircraft mechanic for my entire adult life and been looking for a change for years now and you inspired me to finally pursue Data science. Thank You Ken, I'll be tuning into your channel just to keep my 'ear to the ground' in this exciting field. Also, not sure if you spoke on this before, but what laptop do you recommend for data science? Thanks Buddy and keep up the good work!

[ChocolateMilkCultLeader]

One of the most important things that I wish more people would do is documentation. Of datasets, sources, code, rationale etc. I often have to spend a long time on the dataset exploration phase of the project and trying to figure out things. Everything would be much smoother if people created detailed documents on the different Datasets, how things are collected, what features are used, distributions etc

[DataProfessor]

That’s a great concise video summing up the role of a professional data scientist, this video came at the right moment as I transition from academia to industry and totally relate with many of the points mentioned in the video. As mentioned in the later part of the video, I also find the undesirable aspect of data science to be a double-edged sword as on one end the ambiguity may make things unclear but at the same time also leaves room for experimentation and innovation. Papaya Tracker sounds like an interesting personal project!

[miguelalba]

If you want a role where you want to work only with AI and heavy duty programming go for a machine learning engineer job, if you want to do statistics and sometimes classic ML and analyze data go for a data scientist job.

Working as a ML engineer you spend way more time reviewing and maintaining code (so it's more SE from one side), you use python for model development and c++/java for production (the role is building AI products, so you are all day reviewing, extracting and preprocessing data to then training ML models, evaluate them and match required KPIs)

In the other hand, data scientists positions offer what is described in the video: statistics and data preparation (engineering) and classic ML, there is more flexibility regarding tools and programming languages, and the final purpose is bringing value in a business perspective with the projects you have

For example, In ML engineering you get assigned some set of models that you have to maintain and improve, in data scientist positions you get several projects that constantly change. Both roles have their advantages and interesting scopes

I have been in both, my background is in statistics and I have been working as ML engineer in computer vision in the automotive industry for the last 3 years. Personally, I can say ML engineering is more fun, even more in start ups where you feel you bring a lot of value with your ideas about crazy AI methods

[greatwhitesufi]

I'm working as an Associate Data Scientists at a startup and I honestly haven't needed to code for like 5 months. They've got me focused on an e-learning initiative.

[zuxzux]

This actually answered a couple of questions I had about working professionally. I know that SW Engineers typically work in an agile framework but wasn't sure if Data Science followed that same trend. Thanks Ken!

Now to watch this 400, 000 more times so we can beat Forrest.

[jalenthompson]

I’m 26 and I have one year left in my Data Science curriculum, plus my master’s program. Currently learning MySQL and I’m trying to learn Python and R on the side. It’s really tough since I’m in a data structures course using Java, and that takes up a good bit of my time. Hopefully it all works out for the best.
Also trying to get a good grasp on machine learning and deep learning principles as I go, and making sure I’m sharp with my statistics and linear algebra skills.

[thomasgreen]

Thank you for posting this video. This past year I migrated to a new work role and am listed by my employer as a data scientist. This video confirmed my understanding and also cleared up confusion as many people, my employer, and consumers of our products have different views of a data science work role.

[analisamelojete]

Miss the times when data science was simply called statistics. Now every company wants a “Data Scientist” that can do:

1. Business management.
2. Presentations.
3. Data fetching and collection.
4. Research.
5. Model building and deployment.
6. Model performance evaluation.
7. Cloud environments.
8. Database administration.
9. Pipeline building and management.
10. Client management.

With statistician the responsibilities where quite clear because now every person that works on anything related to data is now a Data Scientist.

[velo]

not all decisions are mad with logic. this hits really hard the long you work in any field :)

[shrutijain]

This video is amazing it literally sums up most of the work data scientist do. I love the first few sec because that's literally me in the morning when I open my mails haha. I think the project management is very up to the points the data scientist role and project manager role has common things is all softs skills and managing work and which is very important in data science job role.
Any video coming up on what data engineer do?

[shrutijain]

I am really liking the Sharpest Mind's concept and something definetly worth looking into

[Ibraheem_ElAnsari]

Thanks for the video Ken ! I can relate to everything you said and also about Asana and the tickets lol .. Hopefully will get used to them.
Glad to see you again !

[fknight]

0:45 Challenge accepted! #TeamSoftwareDevelopment 😄

[osiris]

To help others out and save money I'd recommend do NOT go to college, get a few data related certificates. Start off as a entry level data analyst and while working utilize Udemy and YT on data science info.

After approximately 3-5 years as a data analyst apply to a data science job. Boom! Just saved you money. Any further details needed, just let me know.

Also, Google is your best friend. Learn how to use it being in the data world.

[MrMusicaify]

For anyone curious, the project management framework he’s talking about in the video is “agile”

[cetrick_yeanay]

Thanks Ken! Excited to enter this Domain!

[PlumbobPanda]

11:02 "sometimes kevin doesnt want to use the stupid dashboard that you built and spent 7 weeks on because he doesn't like the color scheme and he wants to use pie charts all over the Had me DYING 💀 I had to pause it. I played this clip for my husband *not a data scientist* and he laughed because THAT is me while I'm working from home.

[mbyoutube]

One of the best reveals I’ve seen about data science. My question now is if it’s better to be a product manager or project manager? Autonomy and authority are some of the most important job factors for me. Are the product and project managers literally managers to data scientists and software engineers?

[pragatitomar]

Finally ...Great insights 🤩

#TeamDataScience