Full-Stack Engineer (AI-Native)

Today we are a large health insurance agency selling coverage nationwide. We have grown through multiple market shifts, from Medicare Advantage to ACA plans, and built a team with many people who have been here for years. That's not an accident, and it's the thing we're most quietly proud of.

We're not here to save the world. Health insurance is a complicated, often frustrating product to sell and service, and we don't pretend otherwise. We care about building something worth showing up to: a company where the work is real, people are treated well, and the clients we serve actually get helped.

We also run a direct-to-consumer subscription e-commerce business in a regulated category. Both businesses have outgrown off-the-shelf tools. As our first full-time engineer, you'll work directly with the founder to build platforms that replace them and turn them into SaaS products we sell to other businesses.

What you will build

• A subscription commerce platform: product catalog, recurring billing with retries and dunning, order routing to fulfillment partners, shipment tracking, a customer self-service portal, and the admin screens our team works in every day.
• A multi-tenant commission dashboard for insurance agencies, sold as a product: carrier commission statements in, clean numbers out for each agency, agent, and downline, covering revenue, chargebacks, persistency, payouts, and forecasts.
• An AI sales assistant that works leads over SMS and turns them into customers: it qualifies, answers questions, handles objections, sends checkout links, and hands off to a human when needed. It plugs into both platforms and is also sold on its own. An early version exists; you will finish it or rebuild it.
• The shared foundation under all three: auth and roles, tenant isolation, background jobs, webhooks, audit logs, and a design system.

What you bring

• 7+ years building production web products, with at least 3 years of senior work before 2023.
• SaaS or e-commerce products you built and shipped: checkout, subscriptions and recurring billing, orders and fulfillment, customer accounts.
• Deep TypeScript, React, and Next.js; Postgres data modeling; typed APIs (tRPC or similar) with schema validation.
• Multi-tenant B2B SaaS: organizations, roles, data isolation enforced on the server, and subscription billing for your own customers.
• Back-end workflow experience: state machines, job queues, retries, webhooks that are safe to receive twice.
• LLM features in production: prompt and tool design, testing conversations before they ship, guardrails, and keeping model costs under control.
• Messaging and CRM integrations: SMS providers, CRMs such as GoHighLevel or HubSpot, and consent and opt-out rules (TCPA, 10DLC).
• Real third-party integrations shipped, including messy data such as CSV or PDF files from partners.
• Money and reporting you trust: you reconcile totals against the source system before you show a number.
• A high bar for UI and UX. You design and build polished, original interfaces yourself: clear hierarchy, considered typography and spacing, every state handled (loading, empty, error), accessible, and fast. You can tell generic AI-generated UI (default gradients, cards on everything, template layouts) from real design, and you do not ship it. Show us a portfolio of interfaces you designed.
• Creative problem solving. When the obvious path is blocked (no API, messy data, a partner that won't cooperate, a budget that doesn't fit), you find a clever, practical way through instead of waiting for perfect conditions.
• Daily, fluent use of AI coding tools (Claude Code, Cursor, or similar). You plan before you prompt, split work across agents by module, keep specs and handoff notes current, review every diff, and verify the running product before you call anything done.
• Comfort with regulated data. HIPAA, PCI, or insurance-industry experience is a strong plus.

How we work

• Small team, direct line to the founder, fast decisions.
• You own architecture, and you ship to production weekly.
• We judge work by the running product, not by tickets closed.
• Design is part of engineering. Our products are used by customers and sold to other businesses, so every screen has to look intentional and trustworthy, never like an AI template.

First 90 days

• Weeks 1 to 2: learn the existing systems, set up the codebase, CI, and the AI workflow (codebase map, handoff notes, acceptance checklists).
• Weeks 3 to 6: commission dashboard MVP live for our own agency, multi-tenant from day one.
• Weeks 6 to 9: AI sales assistant finished and working our own leads, with every conversation logged and reviewable.
• Weeks 9 to 12: the subscription commerce platform running in shadow mode next to the current vendor, with the first product line cut over.