Senior / Staff AI-First Fullstack Engineer

About TFX

We help Americans living abroad file their US taxes — a niche we've been refining for over a decade. Today we serve thousands of expats in 190+ countries and we're scaling fast: double-digit % growth in 2025 and 2026, and a goal to double the business in the next 3–4 years.
Our software is a primary driver of an excellent client experience, and enables growth and efficiency in marketing, sales, and operations. We invest heavily in modern tooling, including AI: the team uses Claude Code, and custom agents as part of the daily workflow, not as a curiosity.

The Role

We're hiring a Senior or Staff fullstack engineer who treats AI as a daily productivity multiplier. You'll ship product features on a mature Rails 8 + Vue 3 codebase, raise the bar of an experienced dev team, and help us extend our internal AI agent infrastructure.
This is an Individual Contributor role — no formal team management, but high technical influence.

What you'll do

Ship product features end-to-end on Rails + Vue — from data model to UI — at a weekly release cadence.
Use AI as your primary way of writing code: decompose tasks so an agent closes them in one or two passes, review AI-generated diffs critically, parallelize across sessions deliberately.
Raise the team's bar through code reviews, agent design patterns, and internal documentation — by example, not by mandate.
Modernize legacy areas of a large codebase — refactor outdated patterns, retire dead code, contribute to ongoing Rails version upgrades.
Extend our internal AI tooling: contribute to the team's library of agents, MCP integrations, and custom workflows.

You're a strong fit if you have

5+ years of commercial fullstack experience (Senior) or 8+ years (Staff).
2+ years of production Ruby on Rails as a primary stack — not side experiments.
Confidence in Vue 3 (Composition API, single-file components) — you can be productive without a long ramp-up.
AI integrated into your daily workflow. You use Claude Code, Cursor, Aider, or Copilot as your primary way of writing code, not as an occasional helper.
Hands-on experience building LLM agents — tool use / function calling, multi-step workflows, output validation, guardrails. Not as a research project — as something you've shipped.
Experience code-reviewing AI-generated pull requests in a team setting — you can tell a good AI-authored diff from a bad one, and give constructive feedback on a colleague's AI output.
A shipping mindset. You can name the last few things you put into production and the impact they had.
High agency and parallel-pathing instincts. You keep multiple threads moving, surface blockers early, and don't wait for a detailed spec to start.
Right to work in your country of residence. We don't sponsor visas.

Nice to have

TypeScript experience in a Vue 3 codebase.
Domain experience in fintech, accounting, or tax.
MCP servers, custom Claude/Cursor skills, or open-source agentic projects — anything that demonstrates you've gone past "user" into "builder" with AI.
Experience modernizing legacy Rails codebases — major version upgrades, large refactors without breaking production.
Observability experience with Sentry, AppSignal, or similar.

What we offer

Fully remote work, any timezone with reasonable overlap with EU and US East.
Competitive USD compensation (discussed on the recruiter screen) — band depends on level.
A mature codebase with real domain depth.
A team that takes AI tooling seriously — not as a gimmick, not as a fad.
A growing, profitable company with a clear runway and a real product market fit.

Hiring process

Recruiter screen (30 min)
Technical conversation with the CTO (90–120 min) — your background, AI workflow, recent shipped work.
Live AI coding session (60–90 min) — a realistic task on a sample repo, using your AI tools of choice.

Apply with resume and interest note to [email protected]
We appreciate your interest and advise that only those selected for an interview will be contacted

Show