Senior Backend Engineer, Agentic AI
Airwallex · San Francisco, CA · 5 days ago
HybridEngineering$150k–$245k/yrFull-time
About the role
This role is based in San Francisco. You’ll design the backbone of our intelligent finance systems: scalable services, resilient APIs, and thoughtful data models. Your work will ensure we can move quickly while maintaining enterprise-grade reliability and security.
What You'll Do
- Design and operate distributed, event-driven systems that process high-volume, high-sensitivity financial data with strict correctness and latency guarantees.
- Build typed, production-grade APIs and services in Python (FastAPI + Pydantic) that power agentic workflows, AI copilots, and real-time financial infrastructure.
- Own data architecture end-to-end, including schema design, event models, ledger primitives, and storage layers optimized for auditability, traceability, and scale.
- Develop resilient infrastructure primitives (queues, retries, idempotency, observability, backpressure handling) that let teams ship fast without sacrificing reliability.
- Work at the intersection of systems + AI, building feedback loops where product data continuously improves models and models autonomously improve system behavior.
- Ship platform foundations, not one-off features internal frameworks, SDKs, and tooling that accelerate every engineer in the org.
- Partner closely with ML + product teams to productionize research-grade ideas into real customer-facing systems.
- Help define the architecture of an AI-native financial stack, not bolting AI onto software, but designing software assuming AI is a first-class runtime.
What You’ll Need To Have
- 5+ years backend experience with Python at scale.
- Deep knowledge of SQL/NoSQL database design.
- Experience with distributed systems, event-driven architectures, or microservices.
- Familiarity with GCP services (Pub/Sub, Cloud Run, Cloud SQL, BigQuery, KMS, etc.).
- Strong grasp of security and compliance best practices.
- Passion for building the backend that powers intelligent products.