6 answers
5
Questions
Asked
803 views
What does an average day look like for a data scientist?
Hi! My name is Jessica. I would like to get a sense of the daily routine of a data scientist to see if this is what I want to do in the future. Thank you! #data-science
Login to comment
6 answers
Updated
Yubing’s Answer
Having skills on both business and technical sides are very important.
As a data scientist in your early career, I would say 40-50% on technical data analysis, programming, and/or statistical modeling work, 20-30% on working with business to understand requirement and documentation, and the rest on empowering yourself through continuous training and learning. While you continue your career on this path, the work might shift more to business side and domain knowledge.
As a data scientist in your early career, I would say 40-50% on technical data analysis, programming, and/or statistical modeling work, 20-30% on working with business to understand requirement and documentation, and the rest on empowering yourself through continuous training and learning. While you continue your career on this path, the work might shift more to business side and domain knowledge.
Thank you, this is really helpful.
Jessica
Updated
Steven’s Answer
Data scientist is a broad term that is used in industry to describe jobs that have vastly different daily routines that will vary from company to company. From my experience and what I have seen based on the job description, the term "data scientist" can be used to label 4 job archetypes. I will try to cover the "daily routine" for each of these job archetypes. The job responsibilities for a "Data Scientist" might cover a single archetype or multiple archetypes depending on the company and team you are applying to.
1. Data Analyst
Focuses on creating reports, dashboards, and insights from data through the use of data visualization and transformations (statistics, etc.). This can be done with programming languages (Python, R, etc.) or through the use of visualization software (Tableau, PowerBI, Excel, etc.).
Daily routines can involve creating new dashboards/reports/insights, generating existing dashboards/reports/insights with the latest data, explaining insights and answering questions about the dashboards/reports/insights, investigating questions or issues around the data and dashboards/reports/insights, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
2. Data Engineer
Focuses on managing data used in reporting and for modeling (training new models and running production models). This can range from infrastructure to the coding logic used for pre-processing the data needed before use. May also be responsible for running existing production models to generate results.
Daily routine can involve creating new pre-processing code for needed data inputs, setting up infrastructure needed to onboard new data pipelines, maintaining existing data pipelines, updating data pipelines for changes in model inputs, running pipelines for generating data, running production models (optional), and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
3. Data Scientist
This is the traditional archetype that most people consider to be the core responsibilities of a data scientist. More applied and responsibilities can be more narrow/focused when compared to a research scientist.
Daily routines can involve developing new models to help solve business problems, maintaining existing models, running existing models, answering questions about model outputs, presenting results and impact of models or model improvements, running analysis/experiments, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
4. Research Scientist
This is role is very similar to a data scientist but focuses more on the open ended or difficult business problems or new modeling approaches to problems that have not been tried before.
Daily routines can involve researching new modeling approaches, setting directions for modeling and/or feature development for existing models, developing new models to help solve business problems, answering questions about model outputs, presenting results and impact of models or model improvements, running analysis/experiments, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
All roles will be a mix of technical and business that boil down to answering questions and making decisions with data, so a big part of the job is communicating the results and impact to people from a variety of backgrounds.
1. Data Analyst
Focuses on creating reports, dashboards, and insights from data through the use of data visualization and transformations (statistics, etc.). This can be done with programming languages (Python, R, etc.) or through the use of visualization software (Tableau, PowerBI, Excel, etc.).
Daily routines can involve creating new dashboards/reports/insights, generating existing dashboards/reports/insights with the latest data, explaining insights and answering questions about the dashboards/reports/insights, investigating questions or issues around the data and dashboards/reports/insights, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
2. Data Engineer
Focuses on managing data used in reporting and for modeling (training new models and running production models). This can range from infrastructure to the coding logic used for pre-processing the data needed before use. May also be responsible for running existing production models to generate results.
Daily routine can involve creating new pre-processing code for needed data inputs, setting up infrastructure needed to onboard new data pipelines, maintaining existing data pipelines, updating data pipelines for changes in model inputs, running pipelines for generating data, running production models (optional), and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
3. Data Scientist
This is the traditional archetype that most people consider to be the core responsibilities of a data scientist. More applied and responsibilities can be more narrow/focused when compared to a research scientist.
Daily routines can involve developing new models to help solve business problems, maintaining existing models, running existing models, answering questions about model outputs, presenting results and impact of models or model improvements, running analysis/experiments, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
4. Research Scientist
This is role is very similar to a data scientist but focuses more on the open ended or difficult business problems or new modeling approaches to problems that have not been tried before.
Daily routines can involve researching new modeling approaches, setting directions for modeling and/or feature development for existing models, developing new models to help solve business problems, answering questions about model outputs, presenting results and impact of models or model improvements, running analysis/experiments, and working with co-workers across the company (business, engineering, and scientist) to help solve business problems.
All roles will be a mix of technical and business that boil down to answering questions and making decisions with data, so a big part of the job is communicating the results and impact to people from a variety of backgrounds.
Thank you so much for breaking it down for me! I have a better understanding of this field now.
Jessica
Updated
Rahul’s Answer
Depends! You could spend your day in lots of meetings, or could do the exact opposite and just spending it by yourself coding up your notebook.
As a DS, you have to wear multiple hats depending on what phase of the current project(s) you're in - gathering requirements, doing feasibility analyses, exploring data to evaluate tasks, fit models to your data, tune them to get best performance, join meetings to demonstrate findings to leaders, and so on.
If you love seeing all of these, it could be the right job for you :)
As a DS, you have to wear multiple hats depending on what phase of the current project(s) you're in - gathering requirements, doing feasibility analyses, exploring data to evaluate tasks, fit models to your data, tune them to get best performance, join meetings to demonstrate findings to leaders, and so on.
If you love seeing all of these, it could be the right job for you :)
Thank you for taking the time to help.
Jessica
Updated
Sulli’s Answer
I'm working as a data scientist and business data consultant in global company and usually work project-based.
In my case, to find out data driven business strategy or digital insights, I have to spend my time for discussion with my team members and also other function teams.
And if I decide what project to deliver, I usually play with my computer and computer program ( :) ).
Find out state-of-art technical methods, check the data analyze methods or modeling accuracy, and review the output with clients and find out insights from the data are main part of delivering projects.
At last, most of the data model and data analytic in business area are focusing on building data-driven strategy.
To make better story, summarizing/making final report is important.
In my case, to find out data driven business strategy or digital insights, I have to spend my time for discussion with my team members and also other function teams.
And if I decide what project to deliver, I usually play with my computer and computer program ( :) ).
Find out state-of-art technical methods, check the data analyze methods or modeling accuracy, and review the output with clients and find out insights from the data are main part of delivering projects.
At last, most of the data model and data analytic in business area are focusing on building data-driven strategy.
To make better story, summarizing/making final report is important.
Updated
Zachary’s Answer
That is a great question! Data scientists are engaged in a wide variety of tasks, and so often every day is unique. I think Yubing had a really good response in terms of the breakdown between the technical/business/training tasks.
Much of the time is spent performing the typical technical tasks that are associated with data science such as data extraction, data cleaning, exploratory data analysis, feature engineering, predictive modeling, optimization, automation, etc.. Often some programming skills are required to complete these tasks, and so having knowledge of SQL and Python/R is very helpful. Programming skills are not necessarily required though, and as time goes on more and more platforms/applications are being created that perform data science tasks without any programming such as Alteryx, DataRobot, H20.ai, etc.
In addition to the technical tasks, there are a significant amount of business related tasks that a data scientist performs. Some examples are gathering business requirements, daily/weekly project update meetings, documenting processes and presenting results. Although it may be tempting to think because of all the technical skills required for data scientists that they would not interact with other people often, but that couldn't be further from the truth. It is certainly a role that requires a significant amount of human interaction, and so developing strong communication skills is essential.
Lastly, since data science is quickly evolving, there will be a fair amount of time dedicated to continued education / training. Data scientists have the opportunity to constantly be learning new analytical techniques and technologies, which in my opinion is a really enjoyable aspect about the job.
Much of the time is spent performing the typical technical tasks that are associated with data science such as data extraction, data cleaning, exploratory data analysis, feature engineering, predictive modeling, optimization, automation, etc.. Often some programming skills are required to complete these tasks, and so having knowledge of SQL and Python/R is very helpful. Programming skills are not necessarily required though, and as time goes on more and more platforms/applications are being created that perform data science tasks without any programming such as Alteryx, DataRobot, H20.ai, etc.
In addition to the technical tasks, there are a significant amount of business related tasks that a data scientist performs. Some examples are gathering business requirements, daily/weekly project update meetings, documenting processes and presenting results. Although it may be tempting to think because of all the technical skills required for data scientists that they would not interact with other people often, but that couldn't be further from the truth. It is certainly a role that requires a significant amount of human interaction, and so developing strong communication skills is essential.
Lastly, since data science is quickly evolving, there will be a fair amount of time dedicated to continued education / training. Data scientists have the opportunity to constantly be learning new analytical techniques and technologies, which in my opinion is a really enjoyable aspect about the job.
Updated
Hang’s Answer
Identifying relevant data sources for business needs
Collecting structured and unstructured data
Sourcing missing data
Organising data into usable formats
Building predictive models
Building machine learning algorithms
Enhancing the data collection process
Processing, cleansing & verifying data
Analyzing data for trends and patterns and finding answers to specific questions
Setting up data infrastructure
Develop, implement and maintain databases
Assess the quality of data and remove or clean data
Generating information and insights from data sets and identifying trends and patterns
Preparing reports for executive and project teams
Create visualizations of data
Collecting structured and unstructured data
Sourcing missing data
Organising data into usable formats
Building predictive models
Building machine learning algorithms
Enhancing the data collection process
Processing, cleansing & verifying data
Analyzing data for trends and patterns and finding answers to specific questions
Setting up data infrastructure
Develop, implement and maintain databases
Assess the quality of data and remove or clean data
Generating information and insights from data sets and identifying trends and patterns
Preparing reports for executive and project teams
Create visualizations of data
This was super helpful, thank you!
Jessica