The Year That Flew By
Introduction
It’s been over a year since my last (and only) post from September 2024. At the time, I was looking for opportunities to continue growing as an engineer. I thought it would be worthwhile to revamp my personal website and start a regular series of posts around both technical and non-technical topics to help crystallize some ideas in my mind.
However, September 2024 was also when I started interviewing for a new job. I spent most of my free time prepping for interviews, without much to spare. After wrapping up the interview circuit, I was thrilled to join Stripe as an engineer in January 2025.
I also got married in 2025, so I was juggling wedding planning alongside the new job, which made for a busy year.
2025 flew by in the blink of an eye, and I’m taking some time now to reflect on the year as a whole.
Personal
Wedding
As I mentioned above, I got married 💍. It was the most significant event of 2025 and of my entire life. My beautiful wife and I had a relatively small wedding at Minnekhada Lodge in August, and we could not have asked for a better day. It was a warm summer day, not too hot, with a small breeze that kept us cool throughout. The venue itself was also beautiful. If you happen to be planning a wedding in Vancouver, B.C., I highly recommend it.
Reading
One hobby I am happy to have picked up is reading! I love the quiet moments in the morning when I’m able to spend some time reading a good book with a cup of coffee. I prefer the morning because my brain feels fried after work and I cannot seem to retain the information.
I’m enjoying the books I’ve been reading, but I’d like to better retain and capture what I learn from them. I plan to add a section about the books I’ve read and share some reflections and key learnings.
Career
I’ve been at Stripe for a little over a year now, and it has been an incredible journey so far. It was exactly what I was looking for in my career - a mix of a high bar for quality and the ambition to move and execute quickly. I am also surrounded by extremely talented teammates whom I feel lucky to be able to learn from every day.
There is a lot I can say about my time here, but I want to distill it down to a few key topics.
- The Greenfield Projects
- The Product Hat
- The Day-to-Day
The Greenfield Projects
In my previous jobs, I never had a chance to design and build a completely new greenfield project. I think this is the norm; the majority of our time as engineers is spent maintaining and extending an existing code base.
It was a new challenge for me to think about how to shape and design something on a blank canvas. In some ways, it’s easier to have the constraints of an existing system and build around them. I find that having this blank canvas can sometimes induce analysis paralysis.
It helped that we have a bias for action and a propensity for “tracer bullet” implementations to kick things off. We also leveraged existing patterns and conventions to set the initial framework.
From my observations, the key was to do deep initial user research, including gathering concrete use cases. From there, it was about understanding the product deeply and shaping the API by obsessing over developer ergonomics. All that being said, it is an iterative process where nothing is truly set in stone.
The Product Hat
One stark difference I noticed was the expectation of engineers. In the past I was able to succeed without deep product knowledge, by focusing primarily on technical expertise. Of course, if you had deep product knowledge, it made you that much better as an engineer. But I don’t think I particularly pushed myself to gain a deep understanding of the product I worked on, and I obsessed over the technical details.
At Stripe, the default operating mode is for every engineer to have a deep understanding of the product. I found that you cannot really have a productive conversation with others if you did not build this foundation. It was a muscle I hadn’t developed, so I struggled at first to keep up.
In hindsight, it seems obvious to me that starting with deep product knowledge should be the first-principles approach, but it was a skill I had to learn.
The Day-to-Day
I was surprised at how much code I was able to produce. It was all the more amplified by the great strides that were made with AI tooling in 2025. I started the year relying on GitHub’s Copilot for autocompletion of code, which I think was pretty common across the industry.
Then we got access to Cursor. I was able to hold a conversation with the AI agent and point it to different places to do refactors, write tests, and implement and complete tasks.
Finally, near the end of the year I picked up Claude Code and that felt like a game changer. I often have multiple machines spun up working on different tasks, all chugging away. Its ability to execute improves month over month. It’s nice to kick off a process before logging off and coming back to see a PR that was created.
The bottleneck now is the review process. I have much more code to review for myself, and in general more code to review from others as well. This shifted my focus from writing code to reasoning about correctness, scope, and long-term maintenance.
[Update 2/28/16] - Since I published this post, Stripe has published their own post about Minions, which is an internal coding agent tool useful for one-shotting a task. I use minions quite extensively, starting out with a long prompt with as much context as possible (providing documentation and code pointers) and steering it along the way. It generally does an amazing job. Minions are already integrated with many of our Stripe internal tooling via MCP.
Onwards and Upwards
This post was a brain dump, and a chance for me to reflect on the year of 2025. A lot has happened, and the way I work has completely changed in the last year (compared to the relatively stable way I worked for the previous few years). I’m expecting more changes ahead, and I’m hoping to document them more intentionally here.
Thanks for reading!