By Trent Brunson, Head of Research & Engineering
Originally published on October 15, 2021
Over the last nine years, I’ve interviewed hundreds of applicants for research and engineering positions. One of my favorite icebreakers is, What kind of project would you choose to work on if you were given a $500,000 budget and one year to work on it with no oversight or consequences? (There’s no wrong answer.) Surprisingly, most people have never indulged themselves in this thought experiment. Why? For one, thinking of a good project isn’t easy!
Trail of Bits engineers are encouraged to dedicate a portion of their workweek to an internal research and development (IRAD) project of their choosing, so they face a similar challenge of having to commit to a project they think they might like. But it’s easy to become rudderless in a sea of security topics, where you may find yourself aimlessly scrolling through HackerNews links, scanning through blogs, and poring over preprints on arXiv.org. So here, I’d like to share a simple exercise that I go through when I’m looking for a new project.
This isn’t about persuading others to buy into your idea or justifying why your project is worthwhile. This is about you, a hard-working and curious soul, discovering topics that you genuinely find interesting and want to pursue—free of judgment, free of consequences.
As you make your decision, think about the factors that may influence your choice:
- What skills you have
- What skills you wish you had
- How much time you have available
- Where you are at in your career
- What it will do for your career
- Whether you will have a team to help
- What impact you are looking to make
Where you are at and where you want to be
Start collecting and organizing information about yourself by making these five lists:
1) Your current skill set. Write down your strengths in areas in which you consider yourself knowledgeable or well read. List broad topic areas to open up the possibility of trying something completely new, or if you want to steer your thinking toward a specific topic area, only include subcategories within a specific domain.
2) What you’re interested in. It’s really simple. What do you think is cool? Maybe you read an article or a blog you thought was clever. Maybe you admired someone’s conference presentation. For this, I prefer to categorize these interests according to how much exposure I’ve had. This way, I can see where I might need to set aside time to learn the basics before making real progress.
3) How long the project will be. This isn’t necessarily meant to be the final date on which you walk away from your work to go do something else. I see this as more of a timeline in which you can stop and ask yourself whether you’re happy continuing or whether you want to choose a different path.
4) How many hours per week you will work on it. This is meant for you to take a look at your current situation and realistically determine your level of dedication. How many hours per week do you see yourself focusing on your project, knowing your schedule, prior commitments, attention span, and ability to work without distractions?
5) Desired outcome. This is meant to tie everything together and ask yourself what it is you want to produce with your effort. The outcome may be subtle, like the satisfaction of learning something new, or ambitious, like publishing a book or writing a dissertation.
Arranging these lists side-by-side helps you see the big picture and discover the different pathways that may lead to a project. I did this for myself to demonstrate how it might look:
The topics in green are ones that I understand fairly well; I could work my way through an academic publication on these topics without much trouble. Those in yellow are ones that I’ve had some exposure to but would need to do some extra Googling and reading to understand some of the subtleties. Those in red are, for the most part, completely uncharted waters for me. They sound cool, but I would have no idea what I would be getting into.
Be resourceful—use mad libs
Reading this chart from left to right, I can begin to think about all the different possibilities.
Using my ____________ skills, I can learn more about ______________ in ___ months if I commit at least ___ hours per week to produce a ______________.
When you start to build statements from your lists, it should become clear what is and isn’t feasible and what you are and aren’t willing to commit to. Here are some examples:
Using my C++ skills, I can learn more about LLVM in 6 months if I commit at least 5 hours per week to produce a peer-reviewed publication.
Sounds nice, but a peer-reviewed publication might be a bit of a stretch. I’ve completed the LLVM Kaleidoscope Tutorial and written some analysis passes before, but I’ve never taken a compiler’s course nor am I familiar with compiler and programming language research. So a blog post or pull request might be more attainable with a 6-month, 120-hour commitment. Also, an LLVM project could be good for my career.
Using my statistics and numerical analysis skills, I can learn more about open-source intelligence in 12 months if I commit at least 8 hours per week to produce a new open-source tool.
I’ve been really interested in Bellingcat’s work ever since I read about how they tracked the downing of flight MH17 over Ukraine to a Russian missile system. Really cool stuff. I think this project and commitment level are similar to a typical IRAD project. At that level of commitment, I would want it to have an impact on my career, so I would need to try and link the project to one of Trail of Bits’ core values. The next step is to narrow the search and see where today’s open-source intelligence tools fall short.
Using my natural language processing skills, I can learn more about topic modeling in 9 months if I commit at least 3 hours per week to produce a blog post.
Three hours per week sounds reasonable for something like this for a personal project. It doesn’t really align with my career goals, but it’s something I’ve read about for the past few years and want to know more about. There are elements of statistics, programming, NLP, and machine learning involved. How cool is that!
At Trail of Bits, I encourage my team to allocate 20% of their workweek to an IRAD project. But when discussing ideas, it’s common to hear people say that they don’t think their idea is good or novel enough, that it isn’t likely to succeed, or that it’s either too ambitious or not ambitious enough.
What makes this exercise for choosing a project so effective is that all the work is simply put into drawing a line from what you do know to what you want to know. Your commitment level is likely to be predetermined by whatever situation you’re in. And the final goal or outcome will be informed by the other four parameters. If done in earnest, this method should produce a whole line-up of possible projects you could find yourself enjoying. I hope you try it, and I hope you find it motivating.
Once again, I’d like to invite our readers to check out our Careers page for all of our current open positions. And if you’re interested in the Senior Software Engineer or Software Security Research Engineer opening, I look forward to hearing more about the IRAD projects you hope to work on at Trail of Bits!