Berkeley Boot Camps

This was a fantastic experience. As a design professional seeking to pivot my existing career into one focused on tech, this course was an eye opening and transformational experience. The skills I learned directly contributed to my being noticed by hiring managers who might have overlooked my resume in the past.  I would highly recommend it for anyone who wants to augment an existing career with web development experience. However, for students who are looking to work as engineers this course is only the first step in a very long journey of consistent disciplined and rewarding learning. 

I benefited a lot from the data analytics and visualization full-time boot camp. I didn't have any programming knowledge and doing this boot camp instilled was a great first step for me in building a new career in data. What the program does exceptionally well in is the career support after the program ends. I dedicated a lot of my time attending career coaching sessions, technical mock interviews, and applying the advice and guidance the career coaches gave me and I managed to land a data analyst role at a start-up in three months. What I found the most difficult for me in the interview process for these companies is definitely the technical interview and the program definitely assisted me with that.

TL;DR: Not worth it for the vast majority of potential students.

An opening disclaimer: UC Berkeley Extension licenses this boot camp from an outside company. So don’t be swayed by the UC Berkeley name re: quality of education. This same curriculum is also used by the University of Denver, Rutgers, and who knows how many other schools. This review is also specifically for the Data Analytics and Visualization boot camp. Their other courses might be better, but I can’t speak to that.

Me:

I had about 4 years experience as a data analyst before starting the class. I was already fairly proficient in Excel and SQL before signing up. I hoped to level up my skillset. My goals were getting more familiar with Python / Pandas (which I’d previously dabbled in), learning best practices around data visualization, and getting some meaningful exposure to ML concepts. I only really accomplished the first.

Instructors:

Julien, Sufyan, and Ahmed were all very helpful and were generally knowledgeable about the material at hand (not always, but no one is an expert at everything). Unfortunately, none of them stayed with Trilogy past my cohort (all of them found better jobs elsewhere). Hopefully other instructors are equally strong. Unfortunately they had some really silly curriculum to work with (see below).

Classmates:

I enjoyed the company of (most of) the other students. I still text with and see them occasionally, and I understand some others see each other pretty regularly. I was a little disappointed with the composition. It was mostly younger folks who were looking to transition into tech/data, rather than more established tech/data professionals looking to level up. The latter group had decent outcomes, the former (much larger) group seems to be struggling 4 months after the class ended (but I haven't done a comprehensive roll call, this is just from LinkedIn monitoring and chatting with friends).

Career services:

Less than useless. All of the UCB Extension boot camps (web dev, full stack, and analytics) are given the same general materials and guidelines, so everyone is kind of corralled down the same path. I attended two video session, scanned their materials, and did a couple resume activities. Maybe helpful for someone fresh out of college, but the material was mind-boggling basic for anyone with any exposure to the white collar employment. Their resume reviewer’s revisions would have made my resume / LinkedIn profile dramatically worse. I’d look like any other “professional” boot camp schmuck with no personality or soul. Basically another sheep blending in with the herd. No thanks.

Curriculum:

Whew boy. The curriculum is a very, very mixed bag. In short, some really great stuff for entry to mid level analytics, some crap that’s indefensibly useless for most data professionals, and too much breadth with not enough depth.

Of the roughly 6 months I was in the program, the first 2.5 months were great for someone looking to break into data (albeit a lot of review for me). Excel and SQL are foundational skills, and exposure to python is a great skill to have.

Then there’s a couple weeks of classes related to HTML/CSS and web scraping. These shouldn’t be top of mind in industry, but I’m glad I learned them. They open up some interesting data sets that I otherwise couldn’t access, which I can deploy for some interesting side projects.

Then there’s a month plus where the class goes off the rails, careening into JavaScript and in particular D3.js. I have a special rant reserved for this later on.

Then the last 2 months of class are a mixed bag. There’s some good information in there (a couple weeks on basic Tableau), a few VERY short sessions on R, then basically a 2 week flyby of advanced Machine-Learning and Big Data so cursory as to be basically useless.

Notice what’s missing here. No meaningful math. No meaningful stats. No meaningful discussion of experimental design and clean data (both technically and just conceptually) best practices. Without these basic skillsets I would NEVER hire someone on my team.

Also notice what’s included. We have both Industry Data 101 tools and concepts (basically beginner Excel) and advanced topics like machine learning. No class should try to cover that much ground in so little time. If someone’s resume hits my desk and their only experience is a Boot Camp Certificate and they claim to know Excel, VBA, Python, SQL, HTML, CSS, MongoDB, JavaScript, D3.js, R, Tableau, SKSci, Hadoop, Spark, and TensorFlow, I am going to chuckle and toss their resume. Those are vastly different technologies, many of which don’t build on or necessarily complement each other, and there’s no way you have a meaningful background in most of them after a few months.

After taking the course, I’m not really sure who it was designed for. It almost seems like a mish-mash of parts from 3 disparate courses:

1.) Basic intro to data analysis and tools.

2.) Front-end development.

3.) Introductory data science for dummies.

The curriculum actually does a pretty good job of (1). After the first couple months, you should have some talking points for an entry-level level analytics gig (maybe associate level if you have a few years relevant experience). But no one hiring entry-level data analysts will expect those same hires to also do machine learning. No one coming out of this class will be remotely qualified for Data Science or ML Engineer jobs unless they already had a quantitative Ph.D or significant substantive technical work experience. Which makes the value of (3) above questionable.

And (2)’s inclusion is just silly. I wasn’t familiar with JavaScript and D3.js beforehand, it’s because they have really limited utility for data analysts. They’re fundamentally web technologies best reserved for front-end data visualization specialists, web developers, or very, very rare edge cases. It makes me wonder if Trilogy just tacked them on because they wanted to reduce their work and recycle material from their Front-end and UI/UX Bootcamps.

D3.js might allow me to work on some interesting pet projects with non-secure data. But incorporating it into a broad intro course for analytics professionals in industry is silly. Most companies will have tools which abstract away the technical aspects of D3.js for their analysts, or it’ll be handled by a team whose job is specifically dealing with web data visualization. Teaching D3.js to an aspiring data analyst is like teaching an aspiring quarterback the nuances of playing wide receiver. Yes, there’s a relationship between the two roles, but most quarterbacks are better served practicing going through their reads and throwing mechanics than catching deep passes or running routes.

Conclusion:

Full disclosure, I did get a job at a large name brand tech company about 3 months into the program. Learning and practicing python from the stronger, early portion of the class had some influence on that. But that skill was more icing on the cake of a strong internal referral, relevant work history, domain experience, and great references. I was also lucky enough to: (1) convince my employer to pay for half of it, (2) already be working as an analytics professional, so I can write it off as an educational expense, and (3) comfortably cover the cost even if 1 and 2 were not true. That is not true of the vast majority of people who took this course, and I suspect most of them feel like they wasted $11,000.