SW Engineer Intern, AI Engineering

About the role 

Join Roku as an Engineering Intern and help build the agentic AI powering America’s #1 TV streaming platform. Roku reaches more than 100 million households worldwide, and we’re building production-grade AI agents across the products and platforms that serve them. We offer a paid 12-week internship, running from June to early September 2027. 

As an Engineering Intern, you’ll be assigned a scoped project with one of our engineering teams and matched to the team that best fits your skills and interests. You’ll work on real problems, such as agents that investigate failures, automate workflows, review or generate code, or help people make complex decisions. You’ll learn how to take an AI-powered idea from prototype to something people actually use, with a mentor alongside you and real ownership from day one. 

Our teams work across a variety of languages and technologies, from user interfaces and embedded software to platform services, automation and developer tooling, and embrace AI-assisted development to boost productivity and quality. 

What you’ll be doing 

• Own a well-defined project that contributes to the success of your team, from defining the problem through to delivery. 

• Build and test AI agents, copilots and automated workflows that perform meaningful, multi-step work. 

• Connect agents to trusted data, services and internal tools, using approaches such as retrieval, tool calling and orchestration, and help define what an agent may do on its own and when it should hand over to a person. 

• Evaluate your work: measure whether it does the job well, check its reliability and failure modes, and keep improving it based on real feedback. 

• Use AI coding tools every day to move faster, and validate everything they produce through testing, review and your own judgment. 

• Write clean, well-tested, documented code, take part in code reviews, and collaborate with engineers, mentors and cross-functional partners. 

Apply if 

• You are a penultimate-year student, or a final-year student planning to go on to a one-year master’s course. 

• You hold a permanent right to work in the UK. Unfortunately, we are unable to consider candidates on a student visa. 

• You are available for full-time work for the whole 12-week program, from June to early September. 

• You are keen to work with AI technologies and have previous experience building AI agents and automated workflows. 

We’re excited if you have 

• Currently enrolled in a degree program in Artificial Intelligence, Computer Science, Engineering or a related field. 

• Proficiency in Python, ideally with projects, open source or hackathon work beyond your coursework. 

• Hands-on experience building an LLM-powered or agentic system that does multi-step work, and ideally that someone other than you has used. 

• Familiarity with agentic patterns such as grounding and retrieval, tool use, orchestration, memory or conversational interfaces. 

• Sound engineering judgment: you check AI-generated work, understand how it can fail, and recognise when simple automation or a human review is the better approach. 

• A solid grounding in computer science fundamentals such as data structures, algorithms, and operating systems or networking concepts. 

• Strong problem-solving and technical communication abilities, and the ability to break down ambiguous problems and deliver in small steps. 

• Proactive mindset with passion for continuous learning and fast iteration. 

How will I use AI at Roku? 

At Roku, we don’t just use AI, we work with it. AI agents and smart tools help power drafts, analysis, and repetitive workflows, while our people bring direction, judgment, and accountability. We’re looking for curious, adaptable builders who can show how they’ve used AI or automation to move faster, raise the bar, and scale their impact. 

We value your AI skills if you have built fluency across the agentic engineering toolchain — coding harnesses like Claude Code or Cursor, MCP servers, custom skills, or agent frameworks.  You know how to drive an agent, verify its output, and ramp up on an unfamiliar codebase with an agent helping you.