Data science bootcamps solve the fundamental problem of training an ever-evolving tech workforce. There is often a romanticization around joining one company early in your career, staying there for several decades, and retiring as a lifelong member of a close-knit team.
Maybe that is how careers used to go. Still, digitization and access to information in this century have normalized the idea of changing not just your employer but your whole career. Specifically, as it relates to tech careers, making a mid-career change has become an increasingly accessible move. It has even led to an industry booming over the last 15 years: the tech bootcamp.
True to its name, bootcamps aim to drastically expedite the kind of experience that you would look for in a master’s degree. This often means that for anywhere from 8 to 24 weeks, you are completely immersed (usually full-time) in a program designed to develop the skills, portfolio, and connections that will maximize your opportunity of successfully pivoting to your new field. That’s not to say that bootcamps don’t present an opportunity to new college graduates, but by and large, the programs are marketed to those looking to level up or completely change the kind of work they’re already doing.
These types of programs are most commonly seen in both front-end, back-end, and full-stack development, as well as data analytics, fintech, and finally, the field we’ll be focusing on in this article—data science.
Data Science Bootcamps: What You Will Learn
As I discussed above, bootcamps are as much about the actual skills and content as they are about professional development and interview preparation. So let’s start with concrete technical skills.
Data science bootcamps tend to operate under the assumption that students have zero experience with programming and working with data. To this end, data science bootcamps will typically start from square one with the basics of Python programming: control flows, defining functions, creating basic algorithms, understanding data structures, and an introduction to standard mathematical/data-oriented packages like NumPy and pandas.
The trademark of a data science bootcamp is the pace, so fitting all of these concepts into a few weeks can be daunting for many students. Still, if the bootcamp is well structured, there should be a lot of support and out-of-class resources for students to talk through problems and reinforce skills. If you’re considering a data science bootcamp, this is key—make sure the bootcamp makes it clear how they will help you succeed. Your understanding of all the material is as much about your hard work as it is about being able to succeed.
In addition to Python, some bootcamps may also choose to teach R programming for data analysis. This can be a great alternative for more statistically-minded students with less programming background. Once students have a fundamental understanding of data analysis in Python/R, the next step is typically to complete a project that allows them to showcase their skills. These projects are essential and possibly the most important part of bootcamp. Projects allow the following:
- Students can start to use code repetitively, allowing skills to become more committed to memory
- For group projects, students can gain exposure to collaborative version control in Github and generally get a feel for what it looks like to work with others on a technical/business-oriented case study. This often involves a realistic simulation of the types of delegation and conflict resolution scenarios that appear in the real world.
- Students can practice presentation and public speaking skills, which at the very least will play a large part in data science interviews, if not the entirety of their post-bootcamp careers.
- Finally, and perhaps most importantly, students can develop a portfolio of real-world questions that they could investigate, analyze, and present using transferable data science skills. For students targeting a specific field, projects can be specifically tailored to allow students an advantage in appealing to certain types of employers.
From here, bootcamps can deviate in many directions depending on the setup of their curriculum. Still, the remainder of the technical skills will often revolve around machine learning and deep learning (sklearn, tensorflow, etc.), specifically related to linear and tree-based models and naive bayes classifiers, support vector machines, neural networks, and k-means clustering.
Data science bootcamps differ heavily with how much they emphasize the mathematical theory behind these models; this can vary from almost zero emphasis on theory, all the way to a complex deep dive into graduate-level equations. Students will be taught how to run the code for these models in their IDE and apply them to real-world scenarios.
Of course, machine learning is nothing without data cleaning and preprocessing, so students will typically get an extensive walkthrough of outlier removal, imputation, working with missing data, feature selection and generation, and other key data cleaning skills. Similar to before, projects will typically accompany these units so that students can showcase machine learning skills to prospective employers in a real-life, value-adding context.
Aside from the above, data science bootcamps may dip into data engineering/architecture or even more business analyst-minded skills like SQL, Tableau, and PowerBI. Every bootcamp makes different decisions around their technical skills curriculum, so make sure to compare them and assess what box you fit into as it regards skills, pace, and learning style.
Data Science Bootcamps for Career Development
While gaining technical skills takes up the bulk of time in a data science bootcamp, many consider the value gained by the career services and professional development that a bootcamp offers. One thing that is typically offered is formal interview preparation, where bootcamp staff will sit down with students to simulate the experience they may have when meeting with real employers.
Similarly, bootcamps may provide a variety of mock technical questions and coding challenges since this assessment style is fairly unique to tech, and most bootcamp participants won’t have seen it before. It’s also very common for bootcamps to assist with resume construction and periodic revisions, appropriately integrating the parts of a student’s former work experience with their newfound technical skills to present most attractively to data science employers.
Given how unique everyone’s background is, this concrete feedback from hiring experts can be extremely valuable and stress-reducing. Alleviating anxiety, in general, is a huge part of working at a data science bootcamp. Any bootcamp educator or alumni is acutely aware of how overwhelming it can be for a student to ingest so much new material in such a short time — material that is supposed to launch an entirely new career. In the end, the bootcamps that offer the most empathetic and personalized experience to students are the ones that will produce the best outcomes.
Perhaps more significant than any of these individual services is the network of employers that many bootcamps aim to connect their students with. Data Science bootcamps will typically have employees dedicated to contacting and fostering a relationship with local companies that, upon students’ graduation, can be connected with the student to set up a prospective interview if it is a mutual match. The reliability and strength of this employer network is often a major factor in a bootcamp’s ranking, as they can substantially expedite the post-graduation hiring search for students.
Bootcamp employer networks generally fluctuate based on location (which is why so many bootcamps are concentrated in the bay area), cost, size, and the age of the bootcamp in general; older bootcamps may have more developed employer relationships. Some bootcamps will offer full or partial refunds for students who do not land a job out of graduation, offering insurance that the bootcamp will successfully facilitate a career change.
Data Science Bootcamps: How Long Do They Take?
Data science bootcamps generally last from 10 to 24 weeks, though the biggest factor guiding this is whether a program is full-time or part-time. Full-time programs are geared toward those making an immediate career change who have taken a long sabbatical from their jobs or have left them entirely. This is more in line with the traditional optics of a bootcamp and will typically be in the 10 to 12-week range.
For students who are continuing to work during their studies, many bootcamps offer part-time alternatives, which generally meet two to three times per week for a few hours over 18 to 24 weeks or more. It’s important to note that bootcamps are specifically designed as an alternative to a master’s program, which could facilitate a similar career change, but over a time frame of 2+ years.
Bootcamps aim to expedite this process by giving you an education that is less in-depth than a master’s program but more centered around hire-ability, networking, and portfolio-building.
Cost of Data Science Bootcamps
At this moment, the current market rate for a reputable full-time bootcamp is around $17,000. Similar to the above, this aims to be a serviceable alternative to a graduate program, which could be double or triple the cost. Some lengthier programs do get as high as $25,000 to $30,000, while some part-time, shorter programs can end up as low as $5,000 to $10,000.
What’s important is assessing the value that a program provides. You can assess if a bootcamp justifies its price tag with a variety of factors, including:
- Access to alumni network and mentors
- Career Services, Resume guidance, etc.
- Quality of instructional staff
- Academic resources, like teaching assistants and tutors
- Employer network
- Hiring guarantee
- Post-graduation hiring events and career fairs
- Emphasis on building your portfolio
Online Data Science Bootcamps
While remote bootcamps already had rising momentum over the last few years, the COVID-19 quarantine pushed remote bootcamps directly into the mainstream. Nearly all programs have quickly pivoted and developed remote alternatives to traditional in-person classes. The move toward online education is expected to continue past the end of lockdown.
Still, in-person bootcamps should also fully recover as the pandemic fades, so prospective students should have two solid options (in-person vs. remote) for nearly every bootcamp that interests them. I recommend in-person programs for indifferent students, as they tend to foster communities, networks, friendships, and general motivation in a more human way. The ease of remote flexibility is much more in-line with what some students require and offers its unique advantages.
To find and compare bootcamps, there are a lot of resources and aggregators on the internet.
It’s important to read through reviews from former students who can outline the program’s specifics and assign a rating for various elements of it.
Just because a program has five stars doesn’t necessarily make it a great bootcamp—it’s important to put that rating in context by also looking at the sample size of reviews (I’d prefer a 4.9-star program with 1,000 reviews over a five-star program with eight reviews).
You can also find valuable reviews by simply searching a bootcamp into Google and browsing through their sidebar on the right; many students choose to give testimonials directly on Google.
Who Should Sign Up?
While data science bootcamps try to market to anyone and everyone looking to pivot into the industry, it’s worth re-emphasizing that these programs are not easy. While students are started from ground zero with programming fundamentals, I’ve seen time and time again (as a former instructor at two different bootcamps) that students with some understanding of programming fair immensely better at these programs than those starting from zero, not just academically, but emotionally, as learning to code from scratch is stressful.
If you’re thinking of joining a data science bootcamp, I recommend you browse YouTube, Coursera, or other online resources for some introductory materials on Python programming. This will allow you to test your level of interest while also in a great position to maximize what you get out of a bootcamp, rather than struggling with the basics throughout. If you know a little Python, have a lot of curiosity and are ready for a quick career change, bootcamps can be a much more efficient and focused way to make the jump than self-learning a full master’s degree.
Do your research, make a lot of comparisons, talk to some recruiters, and if you’re ready to make what is potentially a life-changing decision, do it confidently and get ready for an exciting journey!