Senior AI Platform Engineer — LLM & Agent Infra Lead

Coinbase; Jackson (MS); Hybrid; USD 186,000 - 219,000; Full time

Benefits offered by this job

Equity

Bonus eligibility

Medical insurance

Dental insurance

Vision insurance

401(k)

Job summary

Coinbase is seeking a Senior Software Engineer for the AI Platform to build and run LLM and agent infrastructure used across the company. You will own core platform systems including the LLM Gateway, AI Hub, and agent runtime with sandboxed environments.

You will lead multi-quarter initiatives, collaborate with security, legal, and product, and help scale AI capabilities with responsible use and cost governance. Remote-first with periodic in-person surges to synchronize teams.

Qualifications

• Five years of software engineering experience building ML/AI infrastructure or distributed platform services at scale.
• Experience leading multi-quarter initiatives end-to-end with measurable impact.
• Deep expertise in LLM orchestration patterns and agent frameworks.

Responsibilities

• Own architecture and delivery of core platform systems including LLM Gateway and AI Hub.
• Drive design of Knowledge Base infrastructure connecting data sources.
• Lead AI FinOps capabilities including spend attribution and cost optimization.
• Partner with security, legal, finance, product and cloud providers to ship platform capabilities.
• Build evaluation and observability tooling for production agents.
• Shape applied AI strategy by embedding with teams to build production agents.

Skills

Python

Go

Rust

C

LLM infrastructure

Distributed systems

APIs & microservices

Vector databases

Knowledge bases

Tools

Vector databases

Model serving

Job description

Coinbase is seeking a Senior Software Engineer for the AI Platform to build and run LLM and agent infrastructure used across the company. You will own core platform systems including the LLM Gateway, AI Hub, and agent runtime with sandboxed environments.

You will lead multi-quarter initiatives, collaborate with security, legal, and product, and help scale AI capabilities with responsible use and cost governance. Remote-first with periodic in-person surges to synchronize teams.