Staff AI & Agentic Systems Engineer, Mozilla

• Deep Experience: You have 7+ years of professional software engineering experience, particularly in fast-paced environments where you've shipped complex systems from scratch
• Pragmatic Architecture: A proven ability to judge technical debt. You know how to ship an MVP in days, but you also know how to design a clean API boundary so that early speed doesn't block future scale
• Agentic Mastery: You bring 2+ years of direct, hands-on production experience specifically building autonomous workflows, custom tools, and agentic systems
• LLM Fluency: Deep familiarity with diverse LLMs and orchestration frameworks (e.g., LangChain, LangGraph, CrewAI, AutoGen, or similar tools)
• Startup Mentality: Prior early-stage startup experience, meaning you are comfortable with ambiguity, pivot quickly based on data, and focus heavily on execution
• AI-Native Workflow: Deep, everyday proficiency building software with advanced AI coding tools (such as Cursor, Copilot, or custom LLM extensions) to radically speed up delivery
• Strong Foundations: Comfort writing clean, high-performance code in languages like Python or Go, with a solid grasp of distributed backend systems

Desirable

• Mozilla is building a new browser infrastructure platform for the new generation of web automation. We are looking for an experienced Staff Engineer to join our early-stage team at Mozilla to lead the architectural design and implementation of our next-generation AI agent frameworks and workflows
• This is a "ground-up" role within a fast-moving, nimble team, making it a perfect fit for someone with early-stage startup experience who thrives on a blank slate. You will be the technical anchor for our AI engineering efforts, establishing the patterns for how we build, evaluate, and scale autonomous capabilities
• This role sits within Mozilla’s New Products organization, which operates as an internal incubator for high-potential ventures. You will join a fast-paced autonomous team building a new product from its earliest stages
• Architect Agentic Systems from Zero: Design and deploy production-grade multi-agent frameworks, complex orchestration layers, and custom AI developer tooling
• Balance Speed and Scale: Navigate the tension between rapid prototyping and long-term sustainability, knowing exactly when to cut corners to validate a concept and when to slow down to lay rock-solid architectural foundations
• Orchestrate LLM Workflows: Navigate and integrate various commercial and open-source large language models, selecting and optimizing the right models for specific agentic behaviors
• Accelerate with AI Tooling: Lead by example by extensively leveraging advanced AI coding tools to maximize velocity, while establishing best practices for AI-assisted engineering across the team
• Drive Evaluation and Performance: Implement rigorous frameworks for benchmarking, prompt engineering, and evaluating agent reliability, latency, and accuracy
• Collaborative Technical Leadership: Set engineering standards for code quality and system performance, mentoring other engineers and acting as a primary bridge between research and product