Founding Senior Backend Engineer

Job Description

About Bantyr

\n

\n

Bantyr lets fans pay to talk with the creators they follow. Fans buy messages, media and tickets. Creators earn from every conversation, and the agencies that manage creators run their rosters, chatters and payouts on the same platform.

\n

\n

We run in production on web and iOS. Real money moves through Bantyr every day: purchases come in, revenue is split between creators, agencies and referrers, and payouts go out on a schedule.

\n

\n

The role

\n

\n

You will be one of the first engineers to own Bantyr's backend end to end: the API, the data model, the money paths and the infrastructure underneath them. You report to the CTO and work alongside our chief engineer.

\n

\n

This is a founding seat. You set the standards the next ten engineers inherit: how we design data, test money code, ship migrations safely and respond when production breaks. You will make architecture calls, not just implement them.

\n

\n

You own:

\n

\n

- Correctness of money. Revenue splits, payouts, refunds, chargebacks and disputes are right to the cent and safe to retry.

\n

- Reliability. The API stays fast and up through traffic spikes, deploys and third-party outages.

\n

- Data. A Postgres schema that grows with the product, with migrations that never take production down.

\n

- Security and privacy. Auth, permissions, data retention and erasure are built in, not bolted on.

\n

\n

What you'll work on

\n

\n

These are real problems on our roadmap now:

\n

\n

- Payments and payouts. Harden the pipeline from purchase to creator payout. That covers multi-party revenue splits, scheduled bulk payouts, reconciliation against our payment provider, and money that is idempotent, auditable and never paid twice.

\n

- Risk and disputes. Chargeback and fraud controls: purchase pauses, card blocks, dispute-rate monitoring, and clawbacks that stay consistent under concurrent writes.

\n

- Messaging at scale. Real-time conversations between fans, creators and agency chatters, plus push notifications, background jobs and the queues behind them.

\n

- Agencies and permissions. Multi-tenant access control for agencies, their staff and the creators they manage, including staff acting on a creator's behalf.

\n

- Performance. Find and remove slow queries, unbounded reads and tail latency on an autoscaling API.

\n

- Infrastructure. Containers on AWS, scheduled jobs, observability, and deploys we can roll back with confidence.

\n

- Trust and safety. Conversation review, legal holds, and privacy erasure that holds up to scrutiny.

\n

\n

Our stack

\n

\n

The backend is TypeScript end to end, in a pnpm and Turborepo monorepo shared with our web and mobile apps.

\n

\n

| Layer | What we use |

\n

| API | Node.js, Fastify, TypeScript, Zod |

\n

| Data | PostgreSQL (AWS RDS), Prisma |

\n

| Clients | React web app, React Native (Expo) iOS app |

\n

| Testing and CI | Vitest, Playwright, GitHub Actions |

\n

\n

We use AI coding agents heavily in our workflow. You should be comfortable directing them and reviewing their output critically.

\n

\n

What we're looking for

\n

\n

You have:

\n

\n

- 6+ years building and running backend systems in production, with at least some of that time at an early-stage startup.

\n

- Deep TypeScript/Node.js experience, or deep experience in another backend language and the will to go deep in ours.

\n

- Strong PostgreSQL skills: schema design, indexing, query plans, locking, and zero-downtime migrations on large tables.

\n

- Experience building systems that move money, or that otherwise demand exact correctness: idempotency, reconciliation, audit trails.

\n

- Production ownership: you have been on call, debugged incidents from logs and traces, and fixed the root cause.

\n

- Judgment about tradeoffs. You know when to ship the simple version and when a shortcut will cost us later.

\n

- Clear writing. You can explain a design, a risk or an incident to engineers and non-engineers alike.

\n

\n

Nice to have:

\n

\n

- Payments, marketplaces or payouts (Stripe Connect, Whop, or similar).

\n

- AWS (ECS, RDS, EventBridge) and infrastructure as code.

\n

- Real-time systems, queues and background job design.

\n

- Multi-tenant authorization and security work.

\n

- Experience in the creator economy or consumer social products.

\n

\n

How we work

\n

\n

- Specs before big changes. Significant features start as a written spec covering the design, the risks, the tests and the rollout.

\n

- Small PRs, reviewed and tested. Money and auth code gets regression tests that are proven to fail when the guarded behavior breaks.

\n

- Ship often. We release to production many times a week, behind feature flags where it matters.

\n

- Own it in production. The person who built it watches it ship, and incidents end with a fix to the root cause.

\n

- AI-native. Agents write a lot of first drafts here. Engineers own the design, the review and the result.

\n

\n

Compensation and logistics

\n

\n

| Item | Details |

\n

| Salary | 150,000 to 220,000 USD per year, depending on what you bring to the table |

\n

| Location | Remote. Work from anywhere |

\n

| Employment type | Full-time. We are open to starting part-time while we both assess fit |

\n

| Time off | Flexible time off |

\n

| Equipment | A new Mac |

\n

| Benefits | None yet, since we are an early-stage startup. We plan to add RSUs and health insurance within the next 6 months |

\n

\n

How to apply

\n

\n

Send your resume or LinkedIn to [Kevin to fill: application email or link]. Include a few sentences about the hardest production problem you have owned: what broke, how you found it, and what you changed so it would not happen again.

\n

\n

Our process:

\n

\n

1. Intro call (30 min) with the CTO about your background and what you want next.

\n

2. Technical deep dive (60 min) on a system you built: the design, the tradeoffs, what failed in production.

\n

3. Practical exercise (about 2 hours): a realistic backend problem, such as a money-movement or data-modeling task. AI tools are allowed, and we will discuss how you used them.

\n

4. Team conversations with our chief engineer and a founder.

\n

5. References and offer.

\n

\n

We aim to move from first call to decision within two weeks.

\n