Principal Engineer, Fin AI Agent

What's the opportunity? Fin is the #1 AI agent for customer service, resolving over 1M customer conversations every week. We are on the cutting edge of defining how AI agents should work, inventing new paradigms and discovering new problem spaces. This role is central to the Service Agent pillar, which comprises 12 growing teams building every capability that makes Fin function. You will partner at a strategic pillar level, leveraging broad context across work streams, deep technical understanding of Fin's end-to-end operations, and the autonomy to guide future directions. You will be responsible for initiating new work streams from scratch, including forming new teams, clarifying ambiguous problem spaces, and managing aggressive timelines. Given the intense competitive landscape with fast-moving AI-native startups, we need an individual who thrives under pressure, maintaining technical excellence while operating at startup speed within a well-resourced company with a strong customer base.

What will you do?

• Partner at the strategic pillar level. Utilize broad context across work streams to inform technical strategy and investment priorities, making decisions on focus areas and interventions.
• Spin up 0-to-1 work streams. Assemble new teams, clarify problem spaces, build momentum under aggressive timelines, set high expectations, and drive execution.
• Execute on the most ambiguous, highest-stakes problems. Engage hands-on by writing code and shipping features, maintaining deep involvement in execution.
• Lead experimental work at the AI frontier. Conduct A/B tests, perform prompt engineering, build evaluations, and calibrate accuracy, cost, and latency for LLM-powered features, pioneering new paradigms for customer-facing software.
• Shape long-term technical strategy through execution. Your hands-on work will inform architectural decisions and influence the evolution of how products are built, including data models, system design, and the shift from GUI-first to agent-first interfaces.
• Work across the full stack in an AI-first development environment. Push the boundaries of AI-assisted development using tools like Claude Code with unlimited tokens and a sophisticated ecosystem of MCP servers, skills, and shared context, shaping the engineering organization's workflow.
• Raise the bar for the people around you. Provide direct, actionable feedback to accelerate and develop senior and staff engineers.

What are we looking for?

• The core requirement is a combination of engineering depth and product thinking. We seek someone with profound engineering capabilities coupled with strong product and design instincts. You must be able to both build complex systems and identify the correct problems to solve.
• Operated at real scale and has builder energy. Experience with large-scale distributed systems in production environments, combined with startup DNA: urgency, resourcefulness, and a bias toward building.
• AI fluency. Actively experimenting with AI-assisted development, pushing boundaries, and forming opinions on its impact on software development.
• Deep technical depth with breadth. Ability to go deep into at least one technical domain while easily navigating complex multi-team systems, moving fluidly from business context to technical strategy to implementation details.
• Communication as a superpower. Capability to explain technical investments to leadership, align multiple teams on complex projects, and guide engineers through implementation details.
• Extreme autonomy. Proactively partner with the Engineering Director to identify pillar needs, form convictions, and act on them.
• Critical thinking about the business. Understanding Fin's optimization goals and translating them into technical decisions, including investment, tracking, and defunding work streams.
• At least ~10+ years of experience, with significant time as a technical leader driving complex projects across multiple teams and stakeholders.
• Stack agnostic. While we use Ruby on Rails, React (migrating from Ember), and AWS, what matters most is the ability to learn, build, and operate effectively with AI tools like Claude Code. Proficiency with a high-level programming language and fluency in AI-assisted development will ensure rapid productivity.