Succinct project descriptions are a must to catch the eye of a recruiter
As an up-and-coming data scientist, one of the most common things you’re told is that listing your real-world projects on your resume is a great way to showcase your skills and experience.
However, I can’t be the only one who gets completely stumped when it comes time to write a description that truly showcases a project, not to mention the fact that trends dictate that most resumes should only be one page long.
This means that a formula has to be created that will keep project descriptions short, to the point, and most importantly, effective. With recruiters giving resumes only 7.4 seconds of their time, your project descriptions must catch their eye and make them call you for an interview.
Components of a project description for your data science resume
Recruiters must have a clear idea of the objective or goal of their project.
What were you trying to accomplish with your project? What was your project attempting to resolve? What sort of answers were you looking for? What had to be done within the project? Why were you trying this project?
Now is the time to give a brief overview of your project, including a description of the project objective, timeline, and scope. Keeping in mind that your resume shouldn’t cover more than one page, this section should be brief and focus on clearly describing the items listed above in 1–2 short sentences.
Conducted an analysis to determine air pollution levels in Beijing from July 1st to 15th and how they correlated with hospitalization rates during the same time frame.
The projects you list on your resume don’t necessarily have to be personal ones — in fact, you could list any projects that you were a part of. For example, you could list projects from when you were in college, from previous employment, or from competitions or hackathons.
The main goal is to give the recruiter an idea of what you did and how long it took you to do it.
For a project description where you worked as part of a team, you’ll want to go a step further and indicate your role within the project and your responsibilities. Additionally, for a group project description, now could be a time where you list the technologies or tools you used to fulfill your role within the team.
Example for a personal project: Worked independently for 3 weeks to extract, clean, analyze, and visualize Beijing air pollution-hospitalization rate data.
Example for a group project: Worked 10 hours a week as a data analyst within a team of 4 to produce visualizations of Beijing air pollution-hospitalization rate data using Tableau.
Data: Now is the time to give the recruiter an idea of the approximate data set size and skew. This could involve detailing the exact data set used and providing a link (if available freely online) or describing how and where the data was obtained.
This part of the description could also involve discussing the tools and techniques used to obtain, extract, and clean the data. This section should be 1–2 sentences long.
Example: OpenRefine was used to clean data obtained from Open Dataset A (Beijing air pollution level data between July 1st and July 15th — 350 values) and Open Dataset B (Beijing hospitalization rates between July 1st and July 15th — 50,000 deals)
Models and Tools Used: Arguably the most important section in the project description is your discussion on the models and tools used throughout your project.
This is where you indicate its relevance to the job you’re applying for by focusing on listing the models and technologies used. By listing the technologies that appear in the job ad, you indicate to the recruiter that you’ve successfully used them to complete an analysis. This section should be no more than 1 sentence long.
Example: This project was completed using SQL, Python, Tableau, and a stepwise regression model.
Code: Now is the time to provide a link to the code you’ve written for your project.
The adage always tells us to “show, don’t tell.” If recruiters are intrigued by what you’ve told them about your project, they’ll want you to show them precisely what the bones of your project look like. Therefore, this is the right time to provide a link to your Github or another code-sharing repository. This section should only be one sentence long.
Example: Code for this project can be found here: link
How to put it all together to produce a project description:
Here is what our final project description would look like once you put together all of the individual parts:
Project Description: Conducted an analysis to determine air pollution levels in Beijing from July 1st to 15th and how they correlated with hospitalization rates during the same time frame. Worked independently for 3 weeks to extract, clean, analyze, and visualize Beijing air pollution-hospitalization rate data. OpenRefine was used to clean data obtained from Open Dataset A (Beijing air pollution level data between July 1st and July 15th — 350 values) and Open Dataset B (Beijing hospitalization rates between July 1st and July 15th — 50,000 values). This project was completed using SQL, Python, Tableau, and a stepwise regression model. Results from the analysis found that hospitalizations in Beijing spiked during periods of extreme air pollution, which indicates that hospitals in the city should be prepared for an influx of patients at this time. This study could be improved upon by also taking into account extreme temperatures that may also exacerbate pollution and health issues within the city. Code for this project can be found here: link
- The components of your project description that you need on your resume include the objective/goal of the data analysis, your role in the project, a description of the data you used, a list of the models and tools you used, a link to your code repository, and a short discussion of the analysis results.
- Only include the pertinent details to your analysis to ensure that recruiters are getting the most amount of information in the shortest time possible.
- If you run out of room on your resume, only discuss the project objective, tools, and results — the rest can be discussed in the README file of your code repository.