My 10 Year Journey

Published Jul 17, 2026

This summer marks ten years since I started my first full-time job as a software developer. Here's a quick overview of my career, plus a few tips I'd give the 2016 version of myself.

From Research to Rotations

In the summer of 2015, I completed a Research Experience for Undergraduates (REU) in dynamical systems. The team was fantastic, our results were published, and I had a blast. Afterwards, I had to face the fact that computing was (and still is) my most important interest and hobby. I graduated with a B.S. in Mathematics and Software Engineering and decided to go into industry rather than proceed to graduate school. While research interested me, I preferred the idea of developing a broad skill set rather than becoming an expert in a hyper-specialized domain.

I joined Gartner as part of its IT Rotational Program. The goal was to give new developers the opportunity to work on critical applications across the business, rotating to a new team every six months. I worked on a workflow system for event presentations (C# ASP.NET, React, Knockout.js, OracleDB), a sales enablement tool (Java, React, OracleDB), a contact scoring system for marketing (OracleDB, PL/SQL), and some tooling for internal infrastructure (Terraform, Jenkins, Python). My experience at Gartner particularly shaped how I work with others; rotating across teams gave me access to fantastic mentors, allowed me to work with deep subject matter experts, and taught me to learn as much as possible about business problems rather than particular solutions. I'm happy with the impact I made during my time with each of the teams: some of my code is still in daily use today, apparently!

Going Small

I joined Label Interactive in 2018 because I wanted to gain experience building a wider variety of websites and applications. Although we were only a small design and development shop based in Connecticut, our work spanned from multinational publicly traded companies down to small, local chains. I was one of a handful of developers for the majority of my time there, and as a result I quickly learned to work efficiently and independently. I wore many hats: requirements gathering, system/architecture design, implementation, testing, and monitoring. A significant, thrilling adjustment coming from a large company.

Some of the wackiest highlights of my career took place during these two years. I built a website which was presented to the Dalai Lama (this is not a joke or a typo). I rolled up my sleeves and learned PHP and Drupal to satisfy a client request on a tight deadline. I became development lead on the mobile application for an e-commerce system for a grocery chain, used by real customers to place tens of thousands of orders per month, integrating payments, delivery, and notification services. I automated releases and learned a fair amount of DevOps. Each project's needs were unique and I often had to acquire new skills on tight timelines, but few things compare to the feeling of far exceeding a client's expectations.

Fintech

In May 2020 I joined MayStreet, a market data/infrastructure startup in New York, as a fully remote full-stack developer. MayStreet's core product was ultra-low latency feed handlers (middleware between trading applications and the financial exchanges) written in C++, but we also had a robust data lake of PCAPs representing actual network traffic at the exchanges. I joined as the second member of the team responsible for building customer-facing web applications including a customer portal, a data visualization tool for capture volumes (used for marketing), and the MIDAS system used by the Securities and Exchange Commission.

Within a year I was promoted to Senior Software Developer, then again to VP of Frontend Engineering. While the title change is a fun conversation starter, I was essentially a contributing Engineering Manager for the majority of my tenure. My immediate team primarily worked on MACE, the MayStreet Access Control Engine, and had 4 total developers across North America and Europe. The system, written in Node.js, proved instrumental to our customer portal, Workbench product, and user entitlements on our data lake. Aside from these responsibilities, I reviewed candidate resumes, evaluated technical test results, and conducted interviews to scale the larger customer-facing web application team up to 9 people.

I remained with MayStreet until its acquisition by the London Stock Exchange Group (LSEG) in June 2022. I decided not to continue post-acquisition because I enjoyed the pace of startup life and wanted to join another B or C stage company. My time at MayStreet was very rewarding and I think back on it fondly. It was my first true leadership experience, the problem space was always interesting, and the people were brilliant.

Construction Tech

I worked at OnsiteIQ from September 2022 until February 2025. The core product was essentially Google Street View for construction sites. We dispatched capture specialists to the site (ideally on a weekly basis), processed the video walkthrough footage, then allowed customers to interactively view the 360° imagery. This is particularly useful for progress tracking, claims resolution, and project transparency. The problem space was rich with interesting problems in computer vision, graphics, and data storage.

When I joined, the company had just commenced a rebrand. Since I was the most experienced frontend developer, I volunteered to take charge of the rebrand for the customer-facing product. I built a new design system with Storybook then incrementally reimplemented the site with it. The project was a success in terms of time and budget, and I was proud to have contributed an accessible solution with better performance and a professional aesthetic.

As a result of the project, I was promoted to Engineering Manager (laterally updated later to Staff Software Engineer due to reductions in force, see below). My time was split approximately 50-50 between feature work and team leadership activities during this period. The team delivered new user management and progress tracking features, and everything seemed to be going well until one day in the summer of 2023, we realized that our S3 storage costs were exploding. The cost curve had become a serious problem, and we needed to act quickly.

The root cause was that our processing pipeline was storing data and images inefficiently (and in some cases, redundantly) for customer site walkthroughs. The 360° image viewer was built with these storage assumptions in mind, so it had to be updated as well. While I typically shy away from rewrites, we chose to completely rewrite it given other performance, reliability, and code quality concerns with the existing viewer.

I consider the 360° image viewer rewrite project to be my proudest moment at OnsiteIQ. At the end of the project, we cut our S3 storage and access costs by 78.5%. The rate of storage cost growth became negligible. Professionally, it was also a rare opportunity to use my math degree, in particular vector calculus and linear algebra. I'm immensely thankful that I had this opportunity; I am now more interested and informed about computer vision and graphics as a result.

In spite of the success of the 360° image viewer rewrite project and considerable feature work afterwards, the company faced significant headwinds. I survived several rounds of layoffs, but it was my turn in February 2025. It stung, but it was a necessary choice for the survival of the company. At its height, the engineering team had ten members in total. When I left, there was one full-time developer. As I understand it now, the situation has thankfully stabilized!

Returning to Label Interactive

The next month, Label Interactive asked me to come back and lead engineering. The role is primarily IC work, with some leadership elements. The grocery e-commerce platform I helped build in 2018 is still growing; the system has new layers of complexity that are interesting to account for, and it's rewarding to ship new features and handle incident response for a system customers rely on daily. There are new projects and new clients, and one particular highlight has been building the website for a nationally televised sports broadcast reaching millions of weekly viewers.

The work itself is recognizably the same craft and yet completely different. AI-assisted development is now part of every stage of how my team works: planning, implementation, testing, deployment. I was skeptical for a while but noticed a huge improvement in AI capabilities in 2026. I have opinions about where AI helps and where it confidently makes a mess, but that's out of scope for this post.

Final Thoughts

If I could speak with the version of myself from 2016, here are a few thoughts I would share:

Careers are not necessarily ladders; lateral moves are sometimes the best way forward. Having gone from IC to VP/Engineering Manager then back to IC, I feel that my second IC stretch was better because of the detour. Prior leadership experience has been an incredible asset!
Tools are temporary and workflows change. AI-assisted development has changed full-stack web development completely, and I'm confident that it will continue to change.
Focus on the problem, not the solution. I have seen many expensive mistakes play out for software teams. Focusing on actual user needs rather than particular solutions is critical, especially in the age of AI-assisted development.
The math degree will pay off, just later than you expect. While I never expected that my time spent doing dynamical systems/chaos theory research would directly benefit me outside of academia, it certainly paid off while rebuilding the 360° viewer mentioned above!
Don't stress about never posting on your website 😀