4. Whiteboard programming. This maybe the harderst and more intimidating part of any process. Programming in a blank space. Just you and a piece of paper. Practice this a lot. You don’t need to write the code here perfectly, they want to see you thinking and getting into the solution. Talk and describe your thinking process, don’t be there quite.
5. (Optional) Day of coding in the company. This is the final task, is not that common, but is an invite for their company to be there for a full day, seeing what they do and solving some programming tasks.
Advices for the interview? Here are the ones I could find, they are great:
Brandon Rohrer: https://brohrer.github.io/how_to_interview.html (READ THIS!!!)
The Recruiter side
If you are a recruiter for Data Science positions, first see whom is Data Scientist. Not an easy question but here’s my short answer to that:
A Data Scientist is a person in charge of analyzing business problems and give a structured solution starting by converting this problem into a valid and complete question , then using programming and computational tools develop codes that clean, prepare and analyze the data to then create models and answer the initial question.
What data science is not:
Why is Data Science important?
Data Science and Analytics exists because hidden in the data there are treasures waiting to be discovered.
The Ways a Data Scientist Can Add Value to Business:
This is an extract of an amazing article by Avantika Monnappa
1. Empowering management and officers to make better decisions
2. Directing the actions based on trends which in turn help in defining goals
3. Challenging the staff to adopt best practices and focus on issues that matter.
4. Identifying opportunities
5. Decision making with quantifiable, data-driven evidence.
6. Testing these decisions
7. Identification and refining of target audience
8. Recruiting the right talent for the organization
Do you always need a Data Scientist?
Actually no. I recommend that you read these articles on the subject,
From those, an important quote I can take is:
… leveraging a data science team appropriately requires a certain data maturity and infrastructure in place. You need some basic volume of events, and historical data for a data science team to provide meaningful insights on the future. Ideally your business operates on a model with low latency in signal and high signal to noise ratio.
Without these elements in place, you’ll have a sports car with no fuel. Ask yourself if more traditional roles like data analysts and business intelligence may suffice.
Remember this words: A bad data scientist is way worse than don’t have a data scientist at all.
There’s lot of people wanting a job in Data Science, most of them are really intelligent people, wanting to help and have a path in this area, but be careful before hiring one. I recommend that you search for data science descriptions in the best companies out there, learn about their process, and learn from them.
Also, is not true that they need a PhD to be the best data scientists. They need experience working with data and solving business questions using data science. Before asking for a PhD, ask for knowledge, projects they have worked on, open source projects they built or collaborate, Kaggle kernels they created, related job experience, how did they solve an specific problem.
Data science is not just an IT area, is IT+Business, you need to be sure that the data scientist you hire can adapt to the company, understand the business, have meetings with stakeholders and present their findings in a creative and simple way.
Read this blog post for more information:
and from there some important tips to recruit data scientists:
- Recruiters, work closely with hiring managers to build out accurate job descriptions.
- Iron out nuances to distinguish which types of data scientists will be the best fit for the business’ needs. Hone in on the skillset and experience of the type of data scientist you’re looking for.
- Think long term. Understand how the org plans to leverage this role within the product roadmap.
- Set realistic expectations of available candidate pool. There are more roles than candidates, so recruit accordingly.
- Build a list of ideal candidates and calibrate with hiring manager to gauge fit against reality of talent market.
A good quote on the recruiting process from Vin Vashinta:
Aspiring data scientists want 1 thing from the companies that don’t hire them: an explanation. In many cases their only response is silence. How’s an aspiring data scientist supposed to know what to work on, if companies won’t tell them?
Aspiring data scientists aren’t psychics, but they are hardworking & willing to learn. They’ll rise to the challenge if companies start telling them where the bar is.
Peel back the hiring process at most companies & you’ll find they can’t objectively answer the question, “Why didn’t you interview or hire this person?” I teach clients how much they can learn by examining the candidates they reject as closely as they examine the people they hire.
There’s value to both candidates & employers in the answer to that question. Companies have an opportunity to improve their hiring process. Candidates get the opportunity to be better prepared for their next application with the company.
Beyond the value, it’s the decent thing to do for someone who took the time to apply. Hiring is all about making connections. Silence shows the company doesn’t care enough to treat people the right way. That’s something candidates remember.
