Staff Software Engineer, Backend - Platform (Core AI Automation)

Ready to do the most impactful work of your career?

At Coinbase, we are uncompromising on our mission to increase economic freedom.

The bar is high, the environment is intense, and we like it that way.

This isn't a place for complacency, it’s a place to be pushed past your perceived limits.

If you're ready to build the future of finance alongside people who refuse to settle for "good enough," you belong here.

Coinbase is a remote-first, but not remote-only company.

Expect to get together quarterly for intense in-person working sessions called “surges.” learn more about working at Coinbase.

As a Staff Software Engineer on the Core Automation team within the Platform group, you'll architect and build the Agentic AI systems that are transforming how Coinbase operates.

This team is reimagining customer support and compliance processes for a fully AI-driven world, designing intelligent agents, orchestration frameworks, and measurement systems that deliver delightful customer experiences at scale.

You'll own the technical direction for production AI systems, working across cross-functional teams to bring this vision to reality while building primitives that scale automation across the company.

• Architect and build Agentic AI systems that power Coinbase's compliance automation and other Operations, from intelligent agents through orchestration and guardrails
• Design foundational APIs and measurement frameworks that ensure AI agents are grounded, relevant, and reliably deliver customer delight with minimal hallucination
• Lead technical direction for distributed systems underpinning AI automation, defining architecture patterns and strategic roadmaps in partnership with engineering leadership
• Build reusable primitives and orchestration solutions that enable AI-powered automation to scale across multiple domains beyond the initial customer support and compliance focus
• Mentor engineers on AI system design techniques, coding standards, and production-readiness practices for LLM-based applications