Senior Software Engineer, Payments & Banking
Gravie, Remote
The team
We’re building the payments, banking, and financial accounting platform behind ICHRA NextGen, Gravie’s next-generation Individual Coverage Health Reimbursement Arrangement product. Our team owns obligations, scheduled payments, wallets, ACH processing, banking integrations, and the ledger that tracks every dollar. We are launching a new platform, migrating existing clients onto it, and continuing to ship new functionality at the same time.
We are also building toward automating as much of this work as we responsibly can with AI agents. If you want to apply agentic engineering to a domain where correctness matters and mistakes cost real money, this is that team.
What you will do
- Deliver features end to end through agentic development: write specifications, set guardrails and acceptance criteria, direct agents through multi-step work, validate the results, and own the quality of what ships
- Design and build distributed backend services, event-driven workers, REST APIs, and ledger and relational data models, with idempotency, retries, and failure handling designed in from the start
- Build and maintain integrations with banking and payment partners, including third-party APIs, webhooks, and reconciliation
- Design guardrails for agents and automation that touch money movement, including deterministic checks, approval boundaries, audit trails, and safe failure behavior
- Author and defend Engineering Design Documents that cover both traditional services and agent-driven workflows
- Build the pieces of the client migration to the new platform, such as data validation, backfill, and rollback tooling, while new functionality continues to ship
- Review teammates’ code and designs, mentor other engineers, and share what works in agentic development with the team
- Participate in the team’s on-call rotation, triage production issues, and help drive the follow-up that prevents repeat incidents
What you need
- Production work shipped largely by directing AI agents, and the ability to walk us through how you specified, constrained, and validated it, and what went wrong
- Spec-first practice with explicit guardrails and acceptance criteria, and rigorous review of agent-written code for correctness. For money movement, you know that idempotency, ledger invariants, and reconciliation are where agent output cannot be trusted unreviewed
- Experience building and operating production systems that move money, including payments, ACH, reconciliation, billing, or a ledger or double-entry accounting model
- Five or more years of software engineering experience, including ownership of complex systems from design through production