Staff AI Engineer

About the role

Strava is seeking a Staff AI Engineer to join the GenAI + Discovery Platform team, where you will own the systems that make it easy and reliable for all product teams to ship high value GenAI-powered features at scale. In this role, you will build the shared tooling and platform capabilities that sit at the intersection of AI engineering, platform engineering, and server engineering. You will be responsible for end-to-end delivery of AI capabilities, from architecture and interface design through production deployment and monitoring. This is a high-leverage technical role where you are not just building infrastructure, but building the core understanding of athletes that enables consistent athlete experiences and insight across product surfaces. You will work closely with product engineers, product managers, and data teams to translate cutting-edge AI capabilities into production-ready platforms that facilitate development of AI features providing value to our athletes. The position requires treating AI Platform as a product, bringing engineering rigor such as versioning, contracts, SLAs, monitoring, and deprecation paths to LLM integrations that product teams depend on.

Key facts

What you'll do

• Build for a Well Loved Consumer Product: Work at the intersection of AI and fitness to launch and optimize product experiences that will be used by tens of millions of active people worldwide.
• Build the GenAI + Discovery Platform: Set the vision, design, and own the shared genAI platform including LLM orchestration, prompt management systems, RAG pipelines, search and retrieval services (vector, hybrid and structured search), and evaluation tooling.
• Enable Teams to Ship AI Features Faster: Build self-serve interfaces and golden paths so that product and CUJ engineering teams can build GenAI-powered features without deep AI expertise.
• Own End-to-End AI Capability Delivery: Drive projects from architecture and interface design through production deployment and monitoring, ensuring correctness, latency, reliability, and cost-efficiency of the AI capabilities your platform serves.
• Collaborate Across Engineering and Product: Work closely with engineers and PMs across different verticals to enable new features and build the genAI roadmap, informing product teams on how to consume and leverage AI capabilities effectively.
• Build from a Rich Dataset: Explore and use Strava's extensive unique fitness and geo datasets from millions of users to inform how AI capabilities can extract actionable insights, improve product decisions, and power novel athlete experiences.
• Treat AI Platform as a Product: Bring engineering rigor - versioning, contracts, SLAs, monitoring, and deprecation paths - to AI capabilities and LLM integrations that product teams depend on. You don't ship a prototype; you ship a platform.
• Lead as an Owner: Take end-to-end accountability for the reliability and impact of the systems you build, including their correctness in production, their adoption by downstream teams, and the business outcomes they enable.
• Build for Leverage: Design platforms and tooling that multiply the output of the broader team, reducing the AI infrastructure expertise required for CUJ teams to ship GenAI-powered features.
• Collaborate Across Disciplines: Work fluidly with ML engineers, data engineers, data scientists, and product managers to align on model selection, evaluation standards, prompt strategies, and consumption patterns.
• Raise the Standard: Help establish best practices for GenAI system development, responsible AI patterns, and operational health, and mentor teammates at all levels to do the same.
• Be passionate about the work you are doing and contribute positively to Strava's inclusive and collaborative team culture and values.

Requirements

• 5+ years of experience building and operating complex, production AI or backend systems at scale, with a track record of decomposing large technical problems into well-scoped execution across teams.
• Demonstrated experience building AI platform tooling or developer-facing infrastructure ideally for LLM or ML systems with a strong instinct for API design, versioning, and self-serve patterns.
• Hands-on experience building with large language models in production, including prompt engineering, retrieval augmented generation, and fine-tuning.
• Experience designing and operating reliable, low-latency inference systems and managing the tradeoffs between latency, correctness, and cost.
• Strong understanding of data access patterns, search and retrieval architectures, and evaluation frameworks for AI applications.
• Proven ability to collaborate effectively with cross-functional teams and influence technical direction without direct authority.
• Comfort with ambiguity and a bias for action, balancing rapid experimentation with production-grade reliability and observability.
• Strong written and verbal communication skills to articulate technical tradeoffs to both technical and non-technical stakeholders.

Nice to have

• Experience contributing to open source AI or data infrastructure projects.
• Familiarity with fitness or athlete-related data domains.

Practical notes

This role follows a flexible hybrid model requiring more than half your time on-site in our San Francisco office - three days per week.