The Data Scientist UnicornJoel QuesadaBlockedUnblockFollowFollowingJan 10A Data Scientist Unicorn during his interviewData Science is a trending topic.
Every company is looking for Data Scientists to suddenly become data-driven and leave all those old-fashioned knowledge-based decisions.
Without thinking twice, companies start looking for Data Scientists.
But, what exactly should know a Data Scientist?Data Scientist Job PostingsLet’s check what people are mostly asking for their Data Scientist candidates:PhD degree in a quantitative discipline or equivalent experience (e.
g.
Statistics, Bioinformatics, Computer Science, Mathematics, Physics…)Proficiency with ETL technologies (Talend, Pentaho DI, SQL Server Integration Services).
Strong experience with visualization tools (Tableau, Qlik, Power BI, …)Proficient with Big Data Ecosystems (Spark, Hadoop, Mahout, Kafka, …) and with programming languages: Python, R, Java, C++.
Experience with Machine Learning and Deep Learning Frameworks: TensorFlow, PyTorch, Keras, Caffe, …Here let’s make a pause.
A Data Scientist seems to be someone with a strong background in a quantitative field.
Also, the Data Scientist is proficient with BI tools so will provide awesome dashboards for our business and product guys.
This person will even know finally how to scale our data ecosystem so we do not wait anymore to compute complex queries.
This is becoming amazing.
The Data Scientist will help us implement those top-notch algorithms that everyone is talking about.
Finally, not only Big Data but also Deep Learning will rule our business!After looking at this, there is no doubt.
A Data Scientist can add tremendous value to our company.
Our company is going to become data-driven and Big Data and Deep Learning will drive the company to success.
Ah!.Let’s not forget something very important.
The Data Scientist should be someone who holds great communication skills, business mindset, proactivity and as many soft skills as you can imagine!A Data Scientist at the officeThe disappointmentWe are offering salaries above the mean and we are not able to find any Data Scientist like the one we ask.
What’s going on?.Why do people call themselves Data Scientist if they do not hold a PhD?How is it possible that a Data Scientist is not proficient with complex SQL queries?I can’t believe this candidate has never used Tableau!.How am I going to know the impact of our Proof of Concepts?Narrowing down your searchWell, in the real world, it is very hard to find someone who holds a strong background in quantitative fields but also is proficient with very specific business and data technologies.
You may find those unicorns sometimes but you better narrow down your search based on your problem and your specific needs.
Sometimes you may need a Business Intelligence professional that provides those insights to make you understand better your business and make human data-driven decisions.
Keywords: ETL, SQL, Pentaho, Tableau, Qlik, …In other cases, you may want to find good Data Engineers that scale your data needs.
Those guys that can transform your old data warehouse into something scalable.
Keywords: Kafka, EMR, S3, Dataproc, Dataflow, Redshift, DataLake, Data Warehousing, Java, Scala…Data Science and Data Engineering are relatively new terms and those imply fuzziness about its meaning.
In my opinion, it is better to analyse your specific needs and look for the right candidate or candidates.
Start with something small and iterate through a complete data-driven structure.
But, you are going to need not only Data Scientists but also those guys with strong Databases and Engineering skills.
.. More details