What The F*ck Does A Data Scientist Do?
Each month, let's investigate a career we might know the title of, but have zero idea how they fill their 9-5. This month, let's find out - what the f*ck does a 'Data Scientist' do?
With Alexandre Cohen, Data Scientist
Who are you and what do you do? Legit known you 10 years and nfi what you job is.
Being a data scientist is a lot like being a regular scientist, except we do nothing that scientists do and also never get to hold a glass beaker up to the light with a concocted formula that solves world hunger. So basically it's nothing like being a scientist. Also, it's sort of like creating Artificial Intelligence, but mostly to recognise pictures of dogs and cats, and not at all like in Terminator or Ex Machina.
(See photo below - this is not at all what I do).
Don't get it. Like, okay it's 9.20am, you've strolled in, said hi, had a coffee, had a chat, gone for a walk, it's 10:30am...wait sorry that's my job. You then...
I'm a data science consultant, which means I work with companies and use data science to help them improve their processes. For example, you're a retailer and you want to forecast how much sales you'll have next month so you can know how much inventory you should be ordering, or how much traffic will be in your stores so you can plan how many staff you'll need on a given day and time. And how do factors like weather, holidays, events, etc. affect these? I use mathematical models to find patterns in data that can help make predictions, and businesses use these insights to help make decisions. Plus, I get to wear a suit and use fancy corporate jargon like 'leverage your data', 'upskill your workforce' and 'expense account'.
You sound fancy and adult. Do you like it? What's good and bad about it?
Data science is an amazing field because it combines really awesome cutting edge mathematical modelling with real life uses. With the latest advances in the field, we can even use models to find insights from data in images and/or language. And there's a ton of really cool applications
in the medical field, like building a tool that can find a tumour on a chest X-ray or a brain MRI, or using patient histories to calculate the risk of heart disease.
Sorry, had to correct a few of your Americanisms there ('maths', not 'math'). What kind of person should do it?
I think it takes a good blend of analytical skills and creativity. Obviously, you need fundamentals in math(s) and computer science to understand the mathematical theory behind the models and use programming to apply it. But an important part of the job is also finding creative ways to use these tools in real-world settings, like Business, Medicine, Environment, Sports. That's the part that I find makes it cool and badass (as cool and badass as math(s) and coding can be I guess).