Staff Software Engineer, Genai Platform

Ripple Ripple · Fintech · San Francisco, CA +1 · Engineering

Staff Software Engineer focused on building and scaling production-quality agentic AI systems and the underlying enterprise platform at Ripple, a fintech company. Responsibilities include end-to-end development of agentic systems, defining platform architecture, implementing orchestration patterns, creating data flywheels for quality improvement, and embedding security and guardrails.

What you'd actually do

  1. Develop and deliver production-quality agentic AI systems end-to-end using Python, Go, and/or Java, covering EKS deployment, agent runtimes, memory systems, orchestration, tool integration, and evaluation pipelines that operate across Ripple's polyrepo engineering environment.
  2. Define and advance Ripple's Enterprise Agentic AI and developer platform architecture through practical implementations, reference systems, and production deployments — not abstract diagrams.
  3. Build and implement multi-agent orchestration patterns (planner, executor, reviewer, tool agents) using frameworks such as LangGraph, MCP, Claude Code agent harnesses, or similar orchestration systems, with strong regression coverage and observability.
  4. Run fast, high-quality POCs on emerging agent architectures; harden successful patterns into reusable platform services, APIs, SDKs, and developer templates that engineering teams across Ripple can adopt.
  5. Architect and implement data flywheels that continuously improve agent quality through telemetry, benchmarking, automated evaluation, and structured feedback loops — treating quality, cost, latency, and safety as first-class signals.

Skills

Required

  • Python
  • Go
  • Java
  • Kubernetes (EKS)
  • Service mesh (Istio)
  • Containerized workloads
  • Networking
  • APIs
  • Secure enterprise integration patterns
  • Benchmarking
  • Regression testing
  • Telemetry
  • Observability systems (OpenTelemetry, Prometheus, Grafana)
  • Performance tuning
  • Hybrid environments
  • Managed inference endpoints
  • GPU-based workloads
  • Collaboration skills
  • Cross-functional partnership

Nice to have

  • SDKs
  • APIs
  • Templates
  • Reference implementations
  • CI/CD automation
  • Cross-repo context engines
  • AI-native developer platforms
  • Vector databases (Qdrant)
  • Code embeddings
  • Tree-sitter
  • Retrieval across thousands of repositories
  • Enterprise agentic search and orchestration platforms (Glean, Cursor, Claude Code, Microsoft Copilot Studio)
  • Fine-grained policy enforcement
  • Access controls
  • Sandbox isolation
  • Audit trails
  • Regulated or financial contexts
  • Formal methods (TLA+, model checking)
  • Property-based testing
  • Blockchain
  • Payments infrastructure
  • XRPL/RLUSD ecosystem
  • Open-source contributions

What the JD emphasized

  • production-quality agentic AI systems end-to-end
  • Enterprise Agentic AI and developer platform architecture
  • multi-agent orchestration patterns
  • reusable platform services, APIs, SDKs, and developer templates
  • data flywheels that continuously improve agent quality
  • security, guardrails, sandbox isolation, auditability, and policy enforcement

Other signals

  • Develop and deliver production-quality agentic AI systems end-to-end
  • Define and advance Ripple's Enterprise Agentic AI and developer platform architecture
  • Architect and implement data flywheels that continuously improve agent quality
  • Embed security, guardrails, sandbox isolation, auditability, and policy enforcement directly into agent runtimes