Product Engineer

Full-time · Remote

Gemifi builds the modern equity-compensation and planning platform that financial advisors and RIAs use to guide clients through RSUs, ISOs, and NQOs — exercise timing, growth projections, and tax strategy. We turn messy equity data and dense tax rules into clear, defensible decisions, and we do it as a small, high-agency team that ships fast with AI tools in the loop.

We're hiring a Product Engineer to build customer-facing product end to end across gem-mono, our Bun/TypeScript monorepo: the React/Vite + MUI web app, the Elysia API, our generated SDK, the tax engine, and the agent-ops workflows that keep it all healthy. You should be able to take an advisor's real-world problem and turn it into a shipped, correct feature.

WHAT YOU'LL DO

• Ship features end to end across our Bun, TypeScript, React/MUI, and Elysia codebases

• Own customer-facing surfaces: equity-grant imports, exercise-priority and growth projections, planning workflows, and advisor-facing UX

• Build and extend integrations — custodial/portfolio data (e.g. BridgeFT), AI document parsing (AWS Bedrock + Claude) for PDF/image grant imports, and the ZenStack/Postgres data models underneath

• Use agentic development workflows (Codex, Claude Code, Cursor, Copilot) to plan, implement, test, and harden features

• Help build and improve our internal agent loops for auto-healing, QA, bug-fixing, and incident intake (our agent-ops service)

• Work directly with the CTO and founders to translate advisor workflows into reliable product

• Improve tests, observability (PostHog), review practices, and engineering quality as the team grows

WHAT WE'RE LOOKING FOR

• Strong TypeScript, React, and full-stack product engineering experience

• Comfortable across React/Vite/MUI front ends, typed APIs, and Postgres-backed data models — we use ZenStack + Kysely

• Practical, daily experience with AI coding tools (Codex, Claude Code, Cursor, Copilot, or custom agents) — not just having tried them once

• High ownership: you can break down ambiguous product needs and ship without constant direction

• Strong product taste and real care for correctness, especially around financial data where being wrong is expensive

• Comfortable in a fast-moving monorepo with staging→prod promotion, conventional commits, and a PR/Linear-driven workflow

BONUS POINTS

• Fintech / wealthtech background — equity compensation, RIA or advisor tooling, custodial/brokerage data, or tax

• Financial modeling experience (Black-Scholes, Monte Carlo, AMT/ISO/NQO option-tax math)

• Data import/parsing, automation, or agent-infrastructure experience

• You've built integrations against third-party financial APIs and lived with the messy edge cases

OUR STACK

Bun · TypeScript · React + MUI (Vite) · Elysia · ZenStack + Kysely · Postgres · Pulumi/GCP · AWS Bedrock (Claude) · PostHog · Linear · Codex / Claude Code in the loop

HOW WE WORK

Remote-first, small team, high trust. Specs and Linear tickets are the source of truth, AI agents are part of the everyday workflow, and engineers own features from idea to production. If you like shipping real product, care about getting financial details right, and want to work somewhere agent loops are a first-class part of how software gets built — we'd love to talk.