Staff Software Engineer - AI Copilot

Job Description Job Summary At Zendesk, our focus is helping our customers build great relationships with their customers. Founded by three Danish entrepreneurs, Zendesk has experienced remarkable success and growth while maintaining a fun, positive, and down-to-earth culture. We are looking for a Staff Engineer to join our AI Copilot organisation. AI Copilot is a multi-million ARR product that puts AI directly into the hands of customer service agents and administrators. You will be a technical leader driving the architecture and delivery of our conversational AI backend — the system that enables admins to configure and optimize Zendesk through natural language. We ship to learn: our philosophy is to deliver early, deliver often, and iterate based on real-world customer feedback. What you'll be doing Drive the technical vision and architecture for the Admin Copilot Conversational backend — LLM orchestration, intent detection, context retrieval, and tool execution. Own end-to-end delivery of complex, cross-cutting features — from technical design through implementation, release, and production operation. Be a hands-on technical leader: set engineering standards, make key architecture decisions, and ensure the team builds reliable, scalable systems. Collaborate closely with ML Scientists, Product Management, Design, and partner engineering teams to shape the product roadmap from a technical perspective. Influence technical direction beyond your team — contribute to architecture decisions across the Copilot organisation and adjacent teams. Design and improve evaluation frameworks, guardrails, and safety mechanisms for LLM-powered features in production. Mentor and grow engineers on the team through technical guidance, code review, and pairing. Raise the bar for engineering quality. Identify and address systemic technical risks, performance bottlenecks, and reliability gaps before they become customer-facing problems. What you bring to the role Required 8 years of experience in software engineering, with demonstrated technical leadership in backend systems. Strong experience designing and building distributed systems at scale — APIs, service architectures, and data pipelines. Experience with LLM integration in production systems — orchestration, prompt engineering, evaluation, guardrails, or multi-provider setups. Fluent in at least one backend language (Ruby, Python, Go, or similar); working proficiency across multiple languages. Deep understanding of cloud infrastructure (AWS), containerised deployments (Docker, Kubernetes), and CI/CD practices. A track record of driving technical direction for a team or project — architecture decisions, technical roadmaps, and cross-team alignment. Ability to operate effectively with ambiguity — you can break down complex problems, make pragmatic tradeoffs, and move fast without sacrificing quality. Strong collaboration and communication skills — you influence through technical credibility and clear articulation of tradeoffs, not authority. Preferred Experience building conversational AI or chat-based systems (intent detection, context management, tool use). Experience with ML infrastructure: model serving, inference pipelines, or evaluation frameworks. Familiarity with ML pipeline tooling (e.g., Metaflow) and experiment tracking (e.g., MLflow). Experience with event-driven architectures (e.g., Kafka). Experience with SQL and data infrastructure (Snowflake, dbt). Experience mentoring or technically leading a team of 3 engineers. Tech Stack Our code is written in Ruby and Python Our servers live in AWS Our ML pipelines use Metaflow Our data is stored in S3, RDS MySQL, and Snowflake (with dbt for transformations) Our services and models are deployed to Kubernetes using Docker Heavy usage of LLM technology from multiple providers via our LLM Proxy Our conversational backend uses the Vercel AI SDK for LLM orchestration #LI-AO1 The intelligent heart of custom ... (truncated, view full listing at source)