What you'll do
Brex's AI strategy has three pillars: Product AI builds AI into the products our customers use, Corporate AI gives every employee AI tools to work faster, and Operational AI builds the AI that runs the company. The Operational AI team owns the AI systems and internal platforms that make Brex's operations AI-native, from onboarding and KYC automation to AI-powered customer support. We're a small team with a large surface area, and our work directly shapes how fast Brex can onboard, serve, and grow its customers. As a Senior Software Engineer on the team, you'll focus on Agents Platform, Brex's internal platform for building, evaluating, and running AI agents in production. It already powers critical automations in customer onboarding and AI tools for our sales and customer success teams. You'll own its technical direction, help it evolve from simple workflows into agents that take action, and make it the default way Brex ships AI into its operations.
Where you'll work
This role will be based in our São Paulo office. We are a hybrid environment that combines the energy and connections of being in the office with the benefits and flexibility of working from home. We currently require a minimum of three coordinated days in the office per week: Monday, Wednesday, and Thursday. As a perk, we also have up to four weeks per year of fully remote work.
Responsibilities
Own the technical direction and roadmap of Agents Platform, partnering with Operational AI leadership and the teams that build on it to decide what the platform should do next.
Design and build new platform capabilities, including tool calling, multi-step agent loops, multi-turn and conversational execution, and secure access to internal Brex data and services.
Evolve the evaluation and quality system, from datasets and labeling to run comparison and regression testing, so teams can measure and improve non-deterministic systems with production-grade rigor.
Strengthen the reliability, security, and observability of systems in business-critical flows like customer onboarding, including durable async and batch execution, authorization, cost attribution, and incident response.
Partner with engineering, operations, and go-to-market teams to launch and scale high-impact agents, translating how they work into clear platform requirements.
Drive adoption by making the platform easy to build on through clear contracts, paved roads, documentation, and AI-native developer workflows.
Mentor engineers on the team and across Brex through design reviews, code reviews, and pairing, and help grow the team's technical bench.
Requirements
5+ years of professional software engineering experience, with a track record of designing, building, and operating reliable production backend systems.
Deep, hands-on experience using AI coding tools like Claude Code or Cursor in your daily work, with the judgment to ship faster without lowering the quality bar.
Hands-on experience shipping LLM-powered systems to production, beyond prototypes: orchestrating model calls, enforcing structured outputs, and evaluating quality over time.
Experience with distributed and asynchronous systems and the reliability concerns that come with them, such as retries, idempotency, backpressure, and observability.
A platform mindset: you've built shared infrastructure, developer tooling, or paved roads that other engineers adopt, and you care about developer experience and secure-by-default design.
High ownership and operational excellence, including on-call, incident response, and owning systems end to end.
Strong written and verbal communication, and the judgment to prioritize across many internal customers, including non-engineers. English proficiency is required, as interviews will be conducted in English.
Bonus points
Experience building eval frameworks, LLM gateways, or agent platforms used by other teams.
Experience with Temporal or another durable workflow engine.
Experience with service and agent authorization, such as RBAC, scoped credentials, or agents acting on behalf of users.
Production experience with tool use, MCP, retrieval, or agentic loops.
A background in ML engineering, MLOps, or data infrastructure.
Experience in fintech or another regulated environment, where auditability and model governance matter.
Please be aware, job-seekers may be at risk of targeting by malicious actors looking for personal data. Brex recruiters will only reach out via LinkedIn or email with a brex.com domain. Any outreach claiming to be from Brex via other sources should be ignored.
