Brex
Software Engineer, Forward Deployed Agent Builder
New York, New York, United States$152k–$240kmidAdded 1 month ago
About this role
Brex is seeking a hands-on Software Engineer to design and deploy AI agents that automate internal workflows across the company. You'll embed with teams to understand their processes, then build and ship agentic systems using LLMs and tool frameworks, measuring impact through defined metrics.
What you'll do
- Embed with internal teams to understand workflows and job functions through shadowing and research
- Design and deploy AI agents that automate real workflows across different business functions
- Integrate agents with internal systems, APIs, and data sources
- Define evaluation frameworks and success metrics to measure agent performance
- Build shared tooling and playbooks to accelerate future agent deployments
What they're looking for
- LLM and agent framework development (MCP, function calling, RAG)
- Full-stack AI system building (data, APIs, orchestration, product)
- SQL/NoSQL database design and optimization
- Workflow decomposition and systems thinking
- Cross-functional collaboration and influence
- Production-grade AI/automation systems
- Python or similar backend programming
Benefits
- Hybrid work with 3 coordinated office days per week (Monday, Wednesday, Thursday)
- Up to 4 weeks of fully remote work per year
- Competitive salary: $152,000 - $240,000
- Equity and additional compensation packages available
- Work with cutting-edge AI technology
- Collaborative environment with high autonomy
Opens the official application on the employer’s site. No login required.
Brex
Brex builds mission-critical financial infrastructure and credit decisioning systems that power banking and spending management for companies globally. The company is hiring Backend Software Engineers II to develop scalable systems including underwriting engines, real-time data pipelines, and ML-powered credit models across 200+ markets.
- Website
- brex.com
Likely interview questions
- Walk us through a production AI/automation system you built end-to-end. What was the workflow you automated, and how did you measure success?
- Describe your experience building with LLMs and agent frameworks. What tool-use patterns (MCP, function calling, RAG) have you implemented, and what challenges did you encounter?