The Master Plan

Things are moving quickly here at AQM and a new year brings new challenges. It’s during this time that we have to revisit the entire purpose of AQM’s existence. What are we here to do?

A core focus has always been to teach people how to take away meaningful insight from data. AQM started as a way to teach friends only the useful concepts learned in statistics courses and cut out all the theoretical components that you could easily read up on. Things have changed though; students want more than just lessons. Students, particularly in uni/college, are tired of the classic professor, lecture, assignments format.

An alternative to the classroom is Coursera but even that has become a certification mill where tons of students are trying to ramp up their skills. Coursera is useful but only for a very specific and narrow purpose. So how is AQM any different?

All large scale data analysis/data science is done in teams. There are different roles to be played in actually creating a viable data product. Creating some cool plots and some regression models is not data science which is still a key tool in the process but a small piece of the puzzle. At AQM we focus on creating data products which can be user-focused (think fancy dashboards) or developer focused (for use by other data teams). In order to create data products, you need data teams and that is what AQM really is, a complete data science team.

Since we’re a complete data science team, when you’re a member on a project, your role will be very focused on a specific area of the data product; front-end, back-end, quality assurance, statistical inference, integration and project management. Most large-scale data science initiatives fail primarily because they don’t think about soft skills like project management. Data science isn’t exempt from your normal software engineering practices; it’s just an extension of them.

Usually, you will hold multiple roles but your work will primarily be in one area, depending on your interest, the need of the team and skillset. However, during the first four months of the program, you will go through all the roles in-depth.

So back to the original question; What is AQM here to do? Essentially, we want to achieve the following goals:

Teach data science. This includes all the hard (programming, stats etc.) and soft (project management, collaboration) skills necessary to create a data product. You cannot create a great product without both types of skill-sets.

Build a collaborative environment. We use software like GitHub, Slack and your good ol’ meetings and conversations to emphasize a culture of consistent and constructive communication on various aspects of the program. This is a team environment modeled on how you would work at a firm instead of just learning in a classroom.

Work on awesome projects. We’ve been lucky enough to collaborate with some amazing folks from TransLink and BestBuy on different projects. Our goal is to continue collaboration with them as well as expanding to new and exciting projects with other firms. This allows us to really test the concepts we’re learning against real-world data and is the bread and butter of our program.

Create a community. We’ve engaged quite a few industry folks to help us stay up-to-date on what’s going on in the data science field. Our goal is to hold more talks and hands-on sessions with some great firms in Vancouver. The hope is that AQM is at the fore-front of the drive to create a data science community at UBC, SFU and in Vancouver.