Data jobs, why you shouldn’t care about job titles! 🔍
I recently underwent a job hunt process after an exciting adventure at Riminder. So I was looking for a data scientist position in Paris where I would naturally leverage machine learning knowledge to get insights from data or augment products with smarter features. However, I knew beforehand that I would put some effort in order to discern real data scientist jobs.
This is due mainly to 2 reasons:
1/ Data jobs are relatively new to companies: these jobs appeared gradually in firms since 2013. They are not well defined for many recruiters who don’t understand yet the differences between data jobs (data analyst, data engineer, data scientist). As a consequence, some employers confuse data jobs and label wrongly their job offers.
2/ AI as a marketing tool: In recent years, data science and artificial intelligence has become one of the next big things among others (blockchain, big data, SaaS…). Currently, there is an excitement in this hype around AI and data science. They become tech buzzwords used as marketing arguments to raise funds, sell more or attract talents (look here). So, employers confuses data jobs on purpose in order to hire more. Sadly, in this case, data scientist employees find themselves doing data science stuff rarely. The most striking and frequent example is a data scientist consultant job which turns out to be in many cases a data analyst job (hello consultancy firms, stop doing that please 😉). To be honest, I don’t even understand why employers do this: eventually, employees will be disappointed very quickly and leave.
Obviously, let’s understand the main differences between a data analyst (DA), a data engineer (DE) and a data scientist (DS) jobs.
DA: the data analyst goal is to extract insights from available data. He can explain problems, trends, patterns… He is very analytical and likes to use a lot of dashboards, graphs, charts…
For that, he needs to understand very well the business, can hack his way in some basic computer science stuff and has strong communication skills.
required computer science skills:
- data manipulation (excel, VBA, SQL …)
- plotting packages or softwares (pyplot, seaborn, ggplot, tableau, qlick …)
- results presentation (powerpoint…)
DE: the data engineer worries about cleaning, transforming, enriching, storing and streaming all kind of data that can be useful. As there is more and more data nowadays, this usually involves using big data technologies.
He is very strong in computer science and has to understand the data he is manipulating.
required computer science skills:
- great knowledge about databases (SQL, mongo …)
- big data stacks (hadoop, spark, kafka, elasticsearch …)
- data transformation packages (pandas, nltk, opencv…)
DS: the main focus for a data scientist is how he can exploit the data to build models enabling automatic decision taking when new data is available. That’s machine learning. The most delicate task for a data scientist is to define what the model will be learning and to what extent it will be useful for the product. He often wrangle data to be suitable for his models inputs and outputs and work on models integration, when ready, in production pipeline.
He is very good in maths, good in computer science and has to understand a minimum about the business implications of his work.
required computer science skills:
- machine learning packages (sklearn, h2o, keras, tensorflow, pytorch…)
- data cleaning packages (pandas, nltk, opencv…)
- data engineering and automation (scripting languages)
2 more things:
- data scientist and research scientist/engineer are 2 sides of the same coin. The former is used for applied jobs while the latter for a research position.
- Machine learning engineer is a job that sits between data science and software engineering. Its main preoccupation is to deploy built ML models, and optimize and scale them in a production environnement.
Of course, this kind of separation can be found in an established data team. In small companies or startups, a hybrid data position consisting of the merge of the 3 positions is a common thing.
So now looking at these data scientist job offer examples, here is how I see them:
Job offer titled as data scientist but it’s a data analyst offer actually.
Job offer titled as data scientist but I would guess that it’s data analysis 90% of the time.
Why? Well, they are asking for someone with machine learning skills but when you look at the main activities you can’t help to wonder where will machine learning be applied.
Job offer titled as data scientist but it’s rather a hybrid role of all data jobs.
A genuine data scientist offer.
Another example of a genuine data scientist offer.
That’s it! I hope this article can help candidates decipher easier data job offers.
Last but not least, if you have any comments or critics, please don’t hesitate to share them below. I would be very happy to discuss them with you.
More where this came from
This story is published in Noteworthy, where thousands come every day to learn about the people & ideas shaping the products we love.
Follow our publication to see more product & design stories featured by the Journal team.
