Fullstack Developer (AI)

FULL_TIME

Remote | Semi Senior | Full time | Machine Learning & AI

44 applications Last checked today

LeadsEngine (https://leadsengine.app) is a profitable B2B SaaS that helps Medical supply resellers create, launch, and scale Facebook ad campaigns through a guided 7-step wizard. Customers pay $1,597–$2,497 for access. The product is live, the codebase is solid, and we ship every week.

We are hiring our first engineering teammate. Currently the founder pair-codes with Claude Code, end-to-end. The new hire joins that workflow — not assists it from the side.

We use a subagent-orchestrated workflow modeled on patterns from the Claude Code team at Anthropic. A writing subagent implements features. A separate reviewer subagent reviews the diff. CI runs lint, typecheck, tests, and build before any merge. Humans curate, orchestrate, and judge — we do not type every line by hand. If that workflow sounds exciting, this role is for you.

Our stack: Next.js 14 (App Router), TypeScript, React, Tailwind, Prisma, PostgreSQL, Supabase Auth, Stripe, Meta Marketing API, Vercel.

© Get on Board.

What you'll do

Own features from spec to shipped, working directly with the founder using a subagent-orchestrated workflow.

Roadmap (next 6 months):

• New features across the wizard, dashboard, and admin tools
• Invoice generation, auto-responder, and scheduling integrations
• Location pinning (multi-region campaign targeting)
• Meta ad scaling automation (auto-pause/scale on performance signals)
• Facebook profile scanner (lead qualification at scale)
• Reliability and observability improvements
• New features across the wizard, dashboard, and admin tools
• Invoice generation, auto-responder, and scheduling integrations
• Location pinning (multi-region campaign targeting)
• Meta ad scaling automation (auto-pause/scale on performance signals)
• Facebook profile scanner (lead qualification at scale)
• Reliability and observability improvements

How every feature ships:

1. We agree on the spec — small batches, clear acceptance criteria
2. You drive a writing subagent (Claude Code) to implement + write tests
3. Lint, typecheck, and tests pass locally before push
4. You invoke the reviewer subagent via /review on your own diff — address findings
5. PR with a description that explains why, not just what
6. CI must be green to merge
7. Founder reviews the PR description and spot-checks behavior + security touch-points
8. You merge. Feature reaches customers within days.
9. You drive a writing subagent (Claude Code) to implement + write tests
10. Lint, typecheck, and tests pass locally before push
11. You invoke the reviewer subagent via /review on your own diff — address findings
12. PR with a description that explains why, not just what
13. CI must be green to merge
14. Founder reviews the PR description and spot-checks behavior + security touch-points
15. You merge. Feature reaches customers within days.

You're responsible for orchestrating writing + reviewer agents, judging when the AI is wrong, writing tests that prove behavior, and PR descriptions that survive without you in the room.

What you'll learn:

• AI-first development that ships 3–5x faster than traditional teams
• Production patterns: Stripe webhooks, Supabase Auth, Meta Marketing API
• How to run a small SaaS — direct exposure to product, customers, decisions
• AI-first development that ships 3–5x faster than traditional teams
• Production patterns: Stripe webhooks, Supabase Auth, Meta Marketing API
• How to run a small SaaS — direct exposure to product, customers, decisions

If you've been coding 1–3 years and want to level up fast, this is the right room.

Who we're looking for

A productive, curious, AI-fluent developer who wants to grow with us — not someone who's already done it all.

Core requirements:

• 1–3 years building full-stack apps with TypeScript, React, Next.js (App Router), and a SQL database
• Production experience with Prisma + PostgreSQL — schemas you've designed, migrations you've run
• Daily user of Claude Code, Cursor, or equivalent for at least 6 months. We'll ask how you work.
• Public GitHub or portfolio with recent commits we can read
• Working English — daily async writing + 1–2 live calls per week
• Located in Latin America (Argentina, Colombia, Mexico, Uruguay ideal)
• Treats security, accessibility, and observability as first-class concerns
• Curious and coachable — we're a small team building together. You share your thinking, ask good questions, take feedback well.
• 1–3 years building full-stack apps with TypeScript, React, Next.js (App Router), and a SQL database
• Production experience with Prisma + PostgreSQL — schemas you've designed, migrations you've run
• Daily user of Claude Code, Cursor, or equivalent for at least 6 months. We'll ask how you work.
• Public GitHub or portfolio with recent commits we can read
• Working English — daily async writing + 1–2 live calls per week
• Located in Latin America (Argentina, Colombia, Mexico, Uruguay ideal)
• Treats security, accessibility, and observability as first-class concerns
• Curious and coachable — we're a small team building together. You share your thinking, ask good questions, take feedback well.

How we work (traits that matter):

Skepticism + delegation: trust the process (writing agent + reviewer agent + CI), not any single generation. Claude lies confidently sometimes — you know that.

Reviews behavior, not implementations: you run the feature, read test names, trust the reviewer subagent. No reading working code line-by-line.

Ownership end-to-end: feature, edge cases, error states, reliability — all yours.

Founder-friendly: direct access, doesn't wait to be told.

If you've shipped a feature you're proud of and can explain which parts came from you and which from the AI — talk to us.

Strong plus (if you have it)

Not required — but these help you ramp faster:

AI workflow experience:

• You've used Claude Code's slash commands, skills, or agent features
• You can explain how you validate AI output beyond "I read it" — test patterns, review habits, sanity checks
• You follow AI dev tooling trends and have opinions about Claude Code, Cursor, Windsurf, etc.
• You've used Claude Code's slash commands, skills, or agent features
• You can explain how you validate AI output beyond "I read it" — test patterns, review habits, sanity checks
• You follow AI dev tooling trends and have opinions about Claude Code, Cursor, Windsurf, etc.

Product domain experience:

• Stripe integration (webhooks, idempotency, subscription lifecycle)
• Meta Marketing API (campaign creation, lead generation, audience targeting)
• Supabase Auth (SSR + PKCE flows, session management, RLS policies)
• Vercel deployment and serverless patterns
• Stripe integration (webhooks, idempotency, subscription lifecycle)
• Meta Marketing API (campaign creation, lead generation, audience targeting)
• Supabase Auth (SSR + PKCE flows, session management, RLS policies)
• Vercel deployment and serverless patterns

Background signals:

• A side project that real people use (not just GitHub repos with no users)
• Early-stage startup or solo-founder experience
• You've contributed to or maintained an open-source project
• A side project that real people use (not just GitHub repos with no users)
• Early-stage startup or solo-founder experience
• You've contributed to or maintained an open-source project

Compensation (transparent ladder, performance-gated):

• Start: USD $2,500/mo
• Day 90: $2,500/mo + $1,500 performance bonus
• Day 180: $3,000/mo + $1,500 performance bonus
• Day 270: $3,500/mo + $1,500 performance bonus
• Day 365: $4,000/mo + $2,500 anniversary bonus + $1,500 laptop budget
• Year 2+: Annual review. Strong performers reach $4,500-$5,500/mo.

Plus 13th-month salary every December.

Contract: Indefinite-term, 30-day notice. 10 days PTO + your country's holidays + company shutdown Dec 24–Jan 1.

Tools provided: Claude Max 20x ($200/mo value).

Hiring process:

1. Application screening (we read every one)
2. 20-min async video interview
3. 5-day paid trial — $300 flat to ship a small feature using Claude Code + /review
4. 1-hour live pair-coding session

GETONBRD Job ID: 61320

** Fully remote **

Candidates can reside anywhere in the world.

Показать контакты