About

What I'm looking for

I'm looking for a senior or staff IC role with real ownership: end-to-end product work , full-stack, comfortable with ambiguity. Small team, low process, async-friendly, tenured peers who take the craft seriously. I highly value autonomy and being able to make decisions based on a high-level vision and direction. Ideally, I want to work with people I can exchange knowledge with and grow each other professionally.

After fifteen years mostly on my own, part of what I'm looking for is the thing my solo career has been light on: working on a larger team with other like-minded and skilled individuals.

I'm not very interested in orgs with lots of unnecessary process theater. I prefer smaller, nimble organizations that aren't tied down by bureaucracy and inflexible rules that can't be revisited. I want to work within a system that is critical of itself and demands excellence, openness, and willingness to adapt to changing market/product conditions.

How I think

When handed a problem, my first instinct is to step back and ask what we're actually trying to do, and why. I'd rather spend extra time upfront to clarify the goal than ship something well-built that solves the wrong problem.

I'm not locked into one way of doing things and strongly believe in using the right tool for the job. Being flexible is more important than being comfortable. I like to try out new tools and technologies to evaluate them and see what they offer over what I already know. I'm a big fan of doing less with more.

I prefer an iterative development cycle ("Agile") to tease out exactly how the software should look, work and feel. That said, you only get one "first impression" and quality matters - you will often be judged by how close you are to the target in each cycle and for this reason I think it's worth spending the effort to get as close as you can.

I'm a very detail-oriented person and will sweat the details when it matters. I'm also pragmatic and understand that sometimes the most important factor is speed. I enjoy discussing and weighing tradeoffs before embarking on implementation. At the end of the day, what's most important is solving the problem at hand, both in the immediate sense (this feature needs to work well) and the long-term horizon (this product needs to make sense).

I'm picky about the language I use, in code, writing, and in conversation. I'll often revise an important email or document a few times until I'm satisfied with it, keeping in mind not to "let perfect be the enemy of good."

I enjoy breaking complex ideas or tech down to their constituent components, understanding each piece, and being able to explain everything at any abstraction level with minimal loss of nuance. This has served me well working with a wide range of people - from end-users to CEOs.

On AI tooling

I use Claude and Claude Code extensively for ideas, boilerplate and code analysis these days. My past career trajectory had trended toward deterministic, reproducible, and declarative paradigms (Kubernetes, Terraform/IaC, functional/pure programming), and early AI tooling had felt like a step backwards. Now, with the latest models and harnesses, the tradeoffs have shifted decisively and it has become clear significant time, effort and quality benefits can be had when used correctly.

I've been paying close attention to AI developments and early on had dabbled in personal usage of AI for development using tools like Aider but found them lacking. In 2025, I started using Claude Code professionally here and there - bug fixes, root cause analysis, boilerplate, etc.

In 2026, I came across a perfect opportunity to experiment with and stress-test an AI-First approach to development. I began helping a client with an AI-First SaaS Rewrite, with me in the driver seat but doing very little actual coding and instead focusing on architecture and product quality. The goal was to rewrite an internal legacy PHP app into a modern TS/Node codebase and add on OAuth and Stripe integrations, turning it into a paid SaaS app. I've found AI to be incredibly helpful and a great fit for this use-case - simplifying and porting a legacy codebase to another language and architecture. It's been a great real-world learning experience.

Learnings

I've found a few things are extremely important for maintaining cohesion and continuity while using AI-first or AI-assisted development:

Own the architecture, the abstractions, and the decision making

Rubber-ducking is great but don't let AI make any critical decisions by itself without review - anything you underspecify is liable to be a headache down the road.

Workflow is more important than the model or harness

Research, red/green/blue TDD and bespoke tooling for the agent specific to the codebase (e.g. CLIs, custom skills/commands) are incredibly valuable and worth investing in early.

Hobbies & Fun

I'm an amateur musician - I've been playing guitar since high school and enjoy getting together with other musicians to jam. I love live music and have been to many concerts and festivals over the years (some of my faves - Van Halen, AC/DC, Muse, A7X, Alice in Chains).

On the programming end of things, I'll often whip up a new project to see how far I can take an idea. Games, web scrapers, personal organization tools, visualizers, etc. My local project directory is littered with experiments and throwaway toys, which I use to explore ideas or new tech. I've had a lot of fun tinkering with LLMs and Diffusion models over the last few years - finetuning, ablation, and local models particularly interest me.

I watched some anime as a kid (DBZ, Pokemon, etc) and got into it more after college. Faves: Code Geass, Berserk (yes, even the 2016 adaptation), DBZ (still), Hajime no Ippo, Those Snow White Notes, Rurouni Kenshin, too many to list... I've also found some bangers in various intro/outros I really enjoy: Samurai Champloo ED, Spy × Family S1 ED, Gurren Lagann OP, Classroom of the Elite S2 OP, Dragon Quest: The Adventure of Dai OP1, Jujutsu Kaisen ED1, Jujutsu Kaisen ED2, Berserk 2016 OP1, Berserk 2016 OP2, Hajime no Ippo OP3 (they're all great) to name a few!