Photo by Author | Canva
Data science is crowded in the job market. Employers and recruits are sometimes real a hole that gives you ghosts when you thought you would start talking to your salary.
As if you are fighting, recruiters, and employers are not enough, you also have to fight yourself. Sometimes, success in interviews is really a lack of data scientists. Making mistakes is acceptable. Nothing to learn from them is nothing!
So, let’s separate some of the common mistakes and see how to make them when applying for a data science job.
1. To behave like all the roles
Error: Research from heavy and client positions, until a kick or a timothy challemite is localized, sending each character a resume and cover letter.
Why does it hurt: Because you want a job, not “the best overall candidate for all these positions we are not hiring” the award. Companies want you to fit in a special job.
The software can prefer product analytics a role in startup, while the insurance company is hiring for modeling in R.
Not Your tayloring The CV and the core letter also are at risk of neglecting the CV and the core letter to present yourself as highly suitable for a position.
A fine:
- Read the job details carefully.
- Prepare your CV and Coreletter for job requirements – skills, tools and tasks.
- Not just a list of skills, but also show your experience with relevant applications of these skills.
2. Very common data projects
Error: Collecting data project portfolio with washed projects such as titanic, IRIS datases, MNIST, or home price forecasts.
Why does it hurt: Because the recruiters will fall asleep while reading your request. They have seen the same departments thousands of times. They will ignore you, as this portfolio only shows your business thinking and lack of creativity.
A fine:
- Work with dirty, real -world data. Source plans and data from sites such as stratascratchFor, for, for,. CoglFor, for, for,. Data SFFor, for, for,. Data Hub by NYC Open DataFor, for, for,. Great public datasisEtc.
- Work on less common plans
- Choose projects that show your emotions and solve practical business problems, ideally those that your employer may have.
- Explain the trade office and why your point of view in the business context.
3. Reduce the SQL
Error: Do not exercise enough for SQL, because “it’s easier than a machine or learning”.
Why does it hurt: Because how to know Azigar and avoid more fitting makes you an SQL expert. Oh, yes, SQL has also been very tested, especially for analysts and medium -sized data science roles. Interviews often focus more on the SQL than the Ujjar.
A fine:
- Follow complex SQL concepts: sub -reservoirs, CTEs, window functions, time -to -series, pivoting and repetition questions.
- Use platforms like stratascratch And Late code Follow the real world SQL interview questions.
4. To ignore the thinking of the product
Error: Focusing on the Model Matrix instead of a business price.
Why does it hurt: Because a model that predicts customer with 94 % Rock-akBut most of the flags give consumers who no longer use the product, they have no business cost. You cannot maintain users who have already gone. Your skills are not in space. Employers want you to use these skills to provide value.
A fine:
5. to ignore MLOPs
Error: Focusing on the construction of just one model, ignoring its deployment, monitoring, fine toning, and how it operates in production.
Why does it hurt: Because if you know your model yourself, if it is not useful in production. If you do not know how your model is deployed, re -trained or monitored, most employers will not consider you a serious candidate. You will not do all this yourself. But you have to show some knowledge, as you will work with machine learning engineers to ensure that your model acts in fact.
A fine:
- Understand the three important ways of Data processing: Beach, real time, and hybrid processing.
- Understand Machine Learning PipelinesFor, for, for,. Ci/cdAnd Machine Learning Model Monitoring.
- Exercise by adding workflow design to your projects Data InjectionModel TrainingFor, for, for,. VersioningAnd Service.
- Machine Learning Orchoring Tools familiar with, such as Prefect And Air Flu (For orchestration), Coboflo And Embellishment (For a pipeline summary), and Mlflow And Weight and prejudice (For tracking)
6. not ready for behavioral interview questions
Error: Being unimportant, “Tell me about the challenge facing me”, not to eliminate and prepare questions.
Why does it hurt: These questions are not part of the interview (just) because the interviewer is killing his family life, so he would ask foolish questions in a full office sitting there with you. Testing the questions about how you think and talk.
A fine:
7. Use of Buzz Words without context
Error: Packing your CV with technical and business business, but there is no concrete example.
Why does it hurt: Because “the harmony of the big data has been taken to smooth the AI ​​solution for the expansion of the expansion of productive intelligence from the end to the end,” it really doesn’t mean. You can mistakenly affect anyone. (But don’t trust it.) More frequently, you will be asked to tell what you mean by it and recognizing the danger you do not know who you are talking about.
Correct it:
- Avoid the use of Buzz Words And Talk clearly.
- Learn what you are talking about. If you can’t refrain from using Buzz Words, then add a phrase to each Bazword, which shows how you used it and why.
- Don’t be vague. “I have experience with DL” instead of saying, “I used Long short -term memory Reduce product demand and stockout by 24 % to predict.
Conclusion
It is not difficult to avoid these seven mistakes. It may be expensive to make them, so don’t make them. The process of recruiting data science is complicated and quite terrifying. Try not to make your life even more complicated by suffering from the same stupid mistakes like other data scientists.
Net Razii A data is in a scientist and product strategy. He is also an affiliated professor of Teaching Analytics, and is the founder of Stratskrich, a platform that helps data scientists prepare for his interview with the real questions of high companies. The net carrier writes on the latest trends in the market, gives interview advice, sharing data science projects, and everything covers SQL.