Software Engineer, Early Careers AI/UI

Roku•San Jose, CA

15h•Hybrid

About The Position

Roku is seeking Software Engineers for its Early Careers AI/UI program, Project Ahsoka. This role focuses on building real agentic capability into Roku's platform, which powers over 100 million streaming households. It is primarily a systems engineering job, involving work on telemetry, distributed services, and physical devices. The engineers will develop multi-step agents, detect and fix regressions, build internal tooling, and process real-time telemetry at scale under hardware constraints. The role requires experience with systems or infrastructure, real-time data, distributed systems, platforms, or device work, and a demonstrated ability to build and ship agentic systems that others use. Candidates should be prepared to discuss their projects, accountability for code, and views on the limitations of current agent technology. Fluency with AI tools is an explicit expectation and will be assessed during the interview process, which includes a hands-on engineering exercise on an unfamiliar codebase.

Requirements

Up to 3/4 years of professional engineering experience since completing studies.
Strong systems or infrastructure fundamentals.
Experience with real-time data, telemetry, distributed systems, platforms, or device work.
Demonstrated experience building and shipping something agentic that other people use (work project or serious personal project, multi-step rather than a wrapper around a chat window).
Ability to clearly explain what a project does and what broke.
A clear account of how you stay accountable for code you did not type (e.g., design-first approach, review discipline, personal gates).
A view on where agents are not good enough yet and why.
The existing right to work in the US. Visa sponsorship is not available for this role.

Responsibilities

Develop multi-step agents that perform real work in production.
Detect regressions across services, bisect commits that caused them, and propose fixes.
Generate and review code for merging by other engineers.
Build internal tooling that colleagues will adopt.
Create skills, harnesses, and services to improve team efficiency.
Process real-time telemetry and events at streaming scale, adhering to CPU and memory constraints of home devices.
Address agent problems that arise due to hardware in the loop, such as physical feedback, device fleets, and signals not reproducible on a laptop.