Software Engineer; AI​/Agentic Systems

Requirements

• Production experience building LLM-powered agents — not just prototypes. Must have shipped agentic systems that use tool calling, multi-step orchestration, and human-in-the-loop patterns at scale
• Deep fluency with agentic frameworks and APIs:
  Claude/Bedrock APIs, Amazon Bedrock Agent Core, Lang Chain, n8n, or equivalent orchestration platforms
• Strong experience with event-driven architectures as the execution backbone for agent triggers:
  Apache Kafka, AWS Event Bridge, or equivalent
• Solid understanding of platform-level security for agent systems: scoped IAM/RBAC, execution sandboxing, blast radius containment, and audit trail design
• Experience integrating AI tooling into CI/CD pipelines, developer portals (e.g., Backstage), and software delivery workflows
• Comfortable across the stack:
  Python or Go for backend agent services; familiarity with Git Hub Actions, ArgoCD, and infrastructure-as-code tooling
• Demonstrated ability to evaluate build vs. buy tradeoffs for AI infrastructure and make architecture decisions that balance velocity with long-term maintainability
• 10+ years of software engineering experience, with at least 3 years focused on platform engineering, developer tooling, or applied AI/ML systems
• Track record of leading technically ambiguous, high-impact projects — from architecture through to production adoption — in a large engineering organization
• Strong written and verbal communication skills; ability to translate complex technical concepts for both engineering and leadership audiences
• Experience as a technical anchor on cross-functional teams: comfortable influencing without authority, driving alignment across security, compliance, and product stakeholders
• Prior experience building internal developer platforms, golden-path tooling, or developer experience products is strongly preferred

• We are looking for a Senior Staff Engineer to lead the design, build, and adoption of AI-powered agentic systems across our Developer Experience platform
• This is a highly technical, high-impact role at the intersection of platform engineering and applied AI — you will be responsible for the agentic execution framework that automates delivery, operations, and policy workflows for hundreds of engineers across the organization
• You will own the end-to-end architecture for how AI agents are built, governed, triggered, and observed on our platform — from the event-driven backbone and human-in-the-loop queues through to the agent registry and policy explainability surfaces
• You will also be the internal champion for agentic development practices, partnering with engineering teams to evangelize Claude Code and AI-assisted workflows across the organization
• Architect and build the event-driven execution framework that enables agent-assisted workflows across delivery, operations, and policy systems (e.g., PR automation agents, CI failure triage, onboarding agents)
• Design and implement human-in-the-loop approval queues, error handling, retry logic, and guardrail enforcement to ensure safe, auditable agent execution in a regulated enterprise environment
• Define the execution contract — how agents are invoked, what permissions they hold, how outcomes are recorded — across all platform integrations
• Lead the platform-native integration of Claude Code into scaffold templates, CI failure triage, PR assist workflows, and the developer portal's "what should I work on?" agent
• Build and run an internal evangelism and adoption program for agentic development practices — demos, workshops, documentation, and a developer advocate community
• Partner with engineering teams to identify high-value automation opportunities and convert them into governed, reusable agent workflows
• Design and build a governed catalog of approved agents — metadata, capabilities, trust boundaries, versioning, and usage policies — ensuring agents are discoverable, auditable, and compliant
• Implement lifecycle management for agent versioning and permissions, including deployment pipeline integration, rollback, and decommissioning with full audit trails
• Define and enforce the security posture for agent execution: scoped permissions, least-privilege identity models, execution isolation, and immutable audit logging
• Collaborate with security, compliance, and platform governance teams to ensure all agent execution meets enterprise policy requirements