About the role
Finrep is hiring a hands-on Engineering Lead to own a few of our product lines end-to-end: the architecture behind them, the two to four engineers building them, and the quality of what reaches customers. Our users are SEC reporting leads, controllers, and audit partners. A wrong number can reach a regulatory filing, so every change we ship is checked against source data and the failure cases we already know about.
Our stack is Python and Go services on Kubernetes, a Next.js frontend, and pipelines that move filings, XBRL, and documents at volume. You will make the calls on how those pieces fit together, and you will still be writing code most weeks. You will report directly to the founders, with no layers between you and the decisions. We use AI heavily in our own development. We want someone with a real point of view on where it helps, where it quietly costs you, and what has to be true before an agent's output merges.
Responsibilities:
- Own the architecture of your product lines: service boundaries, the contracts between them, and how they behave under failure.
- Manage two to four engineers: set direction, review code, give feedback that lands, and settle the technical disagreements the team cannot.
- Design the retrieval and agent pipelines behind our reporting workflows and the evals that keep them honest as models and prompts change.
- Make the scaling calls on our filing, XBRL, and document pipelines, where a slow query becomes a support ticket.
- Own production reliability for your services: Kubernetes, CI/CD, instrumentation, incident response, and the fixes that follow.
- Define how the team builds with AI: what agents write, what a human reviews, and what a test must prove before anything merges.
- Hire and ramp engineers as the team grows.
Requirements:
- 4-6 years building production software, including 1-3 years leading or mentoring engineers, formally or not.
- Strong production Python. Go is a plus, or show us a new language you picked up and shipped production code in.
- You have shipped an LLM-backed system that real users depend on and measured whether it got better.
- You have owned a service in production: designed it, ran it, and was on the hook when it broke.
- Prior experience with distributed systems: partial failures, retries, and data consistency.
- Current enough technically to lead a design review.