Before I joined Forward Financing in July for my first-ever co-op, I didn’t really have a clear picture of what kind of work I was going to be doing. I had taken a good number of courses in university, where I learned everything from functional programming to design patterns and AI algorithms. I had done a few small-scale projects in those classes and during my spare time, and even worked a little bit with React and a few APIs to create a web app. Good enough, right? Well, not exactly. I would soon find out that those projects bore little resemblance to what day-to-day software engineering entailed and were nothing like the constantly force-pushed, one-branch repo with commit messages such as “help” or “possible fix”.
The first thing I noticed when I started was the sheer number of repositories that I had to clone and “Oh wow! Look, so many directories and branches” and “Jeez that’s a lot of dependencies” and “What’s a Sidekiq? I thought only superheroes had those.” It took a few weeks to get adjusted into a daily routine after setting everything up, attending meetings to absorb as much knowledge as I could about the different apps and how they worked together.
To be honest, I felt quite unprepared for my role here as a full stack engineer co-op. I didn’t have a baseline at all as to what I could handle, and working with an unfamiliar and huge codebase was daunting to say the least. Not to mention that I had little to no experience working with the tech stack here (Ruby, Elixir, React). What did help was taking on small tasks like improving test coverage and addressing minor tech debt to get exposed to the many different applications the company uses. In addition, everyone on my fantastic team was friendly and eager to help, encouraging me to ask any questions I had. It seemed all dandy and fine, besides the fact that I was still doubtful of my ability to perform well, jumping off the deep end into the world of web development.
Overcoming Impostor Syndrome
Ever since I was first accepted into Northeastern, I felt as if I didn’t really deserve to be at such a good school - all I did in high school was coast by and play lots of video games. I felt the same way every time I was praised for being “smart” on the sole basis that I was majoring in Computer Science, when I got a tutoring job for a class that I really struggled in, and most recently being hired as a co-op here. I wasn’t the kid who asked dozens of questions to the professor out of pure curiosity, nor did I have a particularly easy time in my CS classes, which didn’t help. I had a lot of doubts about myself being competent as a co-op, being unfamiliar with the tech stack and not possessing any prior experience. But as my first co-op comes to a close, I now feel that I’ve learned a substantial amount compared to what I thought I was capable of both in terms of technical and communication skills. And sometimes you do get lucky, but it’s what you make of those opportunities - and I think that I’ve done a satisfactory job.
Really, You’re Not Bothering Anyone
If there’s one thing I could change from my time working here, I would have asked a lot more questions. I thought my questions were usually basic and could be easily found searching online, so I would always do that before asking any. The few questions I did ask had very simple answers, so having a habit of doing things by myself, I felt that I hadn’t tried enough to answer the question myself. However, I would end up not being able to find the answers I was looking for and kept feeling like a burden asking simple questions, like what a specific query did or what the stack trace from the error meant.
I would have asked so many questions about how the apps work from a business perspective by shadowing the stakeholders (underwriters, prequal analysts) and seeing the overall workflow and what issues there were. I only got a good feel for one piece of the underwriting app - our internal risk analysis tool for funding merchants - after struggling on a card (described in detail later) and asking lots of questions directly to an underwriter. As weeks passed by and I got more comfortable at the job, I got better about leaving aside that feeling and asking the question anyways - but I do wish that I had done it sooner.
If I had to do it all over again, I would’ve asked about five times more questions than I did regardless of how basic or obvious the answers seemed. Funny thing is, I’ve heard this advice over and over again and yet it fell on my deaf ears until I went through the experience myself. But I know now for the future that I might end up a bit annoying sometimes but at least I’ll learn a lot!
Always in Flux
A rather unique experience I had during my time at Forward Financing was quickly adjusting to constant change. From when I joined in July to a couple months later in December, the company went through a few big changes. We were, and still are, in the process of hiring several product managers, which taught me how important they are in keeping the software development lifecycle (SDLC) running smoothly especially with scoping out tasks and minimizing the friction between stakeholders and engineers.
I would like to start by talking about a rather significant change in our engineering process where our sprint of three weeks plus a few days reserved for what we call slack time - a short period where the team works on tech debt and projects that are not of immediate business concern - was changed to two weeks with no slack time. However, I feel that this did not pan out too well due to frequent meetings, causing a lack of productivity. Additionally, tech debt piled up due to immediate business needs being prioritized.
These are just some of the observations I made during this slightly chaotic time; others might have completely different interpretations and thoughts of the matter. As of the time of writing this post, we have reverted back to a three-week sprint with a week’s worth of slack time. From this, I learned that processes can and will change, but they don’t necessarily need to be permanent and should be critically assessed to see if they work out or not. The same process that did not fit into our workflow may be the best one for another engineering team.
Bite off a Bit More than You Can Chew
I had absolutely no idea how much I could take on in terms of workload and deliver in a timely manner. What if I didn’t finish my cards for the sprint? What if I break other features with my new one? But with time I’ve come to realize that it wasn’t about delivering everything on time or having a perfect feature, but rather to extend myself a bit more every time and fail often. During one sprint I had just one card and was hesitant to ask for more, even though they weren’t big tasks, but I took another and it went just fine. And although it might not work out everytime, failure in software engineering is not such a huge issue. Just as with tests, things just might not work the first time around and that’s okay.
Hash#Dig then Dig a Bit More… Even If You Find a
So during my last few weeks here, I decided to pick up an seemingly straightforward 3-point task that involved pulling date info from our background check vendor and our preferred credit bureau to compare with data from a merchant’s application. All I was actually doing was writing a small service to grab the data and load it in the ERB template for the view. Little did I know how awful APIs could be, as well as that it would take me a month to complete this tiny feature. I would say that I spent maybe a couple of days actually writing the service and helper to create the view, but pulling the data was a nightmare. Having never worked with the Underwriting App before, diving deep into a binary payload and digging out 10-times deep nested hashes for one Date object was not the most fun thing I had in mind.
In the beginning I thought it was a one-hour task, as our Document models had methods that pulled dates from the background check vendor and the credit bureau.
However, I soon found out after that those were the dates of when the data was pulled, not the actual
dates that I needed. Figuring out the specific document type from the vendors that contained the data was a hassle, not to
mention that I probably must have seen more than a thousand
nil values helplessly trying to dig for the proper
object. After wrestling with the task for a week, I figured out the exact method to pull the data,
but it was not pulling the correct data from some applications. Then another week of struggle passed until I realized that the backend logic was already done in a separate part of the app - until that point, I didn’t know I could have felt this demoralized and plain dumb for not seeing it sooner.
However, I found a silver lining after going through this process: I had dug a lot deeper into the app than I would have if I were to find the preexisting function initially, getting quite familiar with the app’s object relations, MVC design, as well as discovering the nuances of Ruby. I would do it better the second time around, but that’s exactly what experience is for.
end… Or Not
And with less than a few days left of my first co-op, I would like to thank the entire engineering team for making my time here wonderful and memorable. We have a great team here not only in Boston, but all over the world in the UK, India, Dominican Republic, Brazil, and Paraguay - all of whom flew in just last week to make for a memorable holiday season and end of the year. The clichéd response to “What’s your favorite part of the job” is “The people” in every interview I’ve had, but for this co-op it had substance and proved itself well. I have had a lot of exciting and unique experiences here that I would be hard pressed to find at another company and am going to take the lessons learned here forward (pun absolutely intended) into the future.
I would like to thank everyone here at Forward Financing and shoutout a few people for making my time here so very pleasant:
Kelvin - For sound mentorship/guidance and establishing #BestRow
Zach - For solid leadership and nuanced humor (and being the only other person I know to use Micro)
Scott - For bringing so much energy into the office everyday
Natacha - For helping me find footing by pair programming extensively when I first started
Mike - For diving deep into seemingly infinitely nested Hash structs with me
Nate - For the thorough code reviews and red-hot enthusiasm for programming languages