Campfire
AI Engineer - Agents
About this role
Campfire seeks a Staff AI Engineer to design and operate production AI agents that automate accounting workflows for startups and mid-market tech companies. You'll lead agent architecture development, ship customer-facing features, and establish reliability practices while balancing software engineering excellence with hands-on AI implementation.
What you'll do
- Design and build production AI agents that perform real accounting tasks using tool calling, orchestration, and data retrieval
- Lead agent architecture decisions and establish engineering practices that ensure reliability and quality in production
- Own agent quality through evaluations, tracing, monitoring, and systematic debugging of production failures
- Collaborate with product and design to translate ambiguous requirements into trustworthy customer-facing features
- Mentor engineers and raise engineering standards while remaining actively involved in coding
- Balance tradeoffs between accuracy, latency, cost, and complexity to choose appropriate solutions
What they're looking for
- Python and backend systems design
- LLM-powered agent development and deployment
- Distributed systems and data modeling
- Production monitoring, evaluation, and observability
- Agent permissions, data access control, and failure recovery
- Staff-level technical leadership and mentorship
- API design and testing practices
- Financial workflows and domain understanding (plus)
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Campfire
Campfire builds accounting software designed for startups and mid-market tech companies, with a focus on automating financial workflows through intelligent features. The company is hiring AI engineers and full-stack engineers to develop and deploy machine learning solutions and modern platform capabilities in their San Francisco office.
View all jobs at CampfireLikely interview questions
- Can you walk us through a production AI agent system you built? What were the key reliability and quality challenges you faced?
- How do you approach evaluating and monitoring agent performance in production, especially around accuracy and failure modes?