Sr. SW Engineer, Machine Learning

About Roku

Roku is the leading TV streaming platform in the US, with a mission to connect the entire TV ecosystem by helping consumers discover content, enabling publishers to grow audiences, and giving advertisers powerful ways to engage viewers.

About the Team

The Ad Engineering team builds the platform behind Roku's advertising business. Within Ad Engineering, the Business Applications team owns the systems and tooling that power the advertising revenue lifecycle. The platform internal teams and advertisers rely on to plan, price, book, and grow campaigns at scale.

About the Role

We are hiring a Senior Engineer to lead the architecture and implementation of a next-generation, agent-native, AI-powered business applications platform. This is a high-impact role focused on building intelligent, recommendation-driven experiences that make day-to-day revenue decisions with pricing, booking, upsells, retention, and media planning faster, smarter, and more accessible to the teams that run Roku's advertising business.

You will architect and build the end-to-end system for AI-driven recommendations and decisioning, from business and customer signals through to actionable, validated outputs. You will work closely with ML, backend, frontend, data, and business stakeholders to deliver a production-ready platform that improves speed, accuracy, and scalability across the advertising revenue workflow.

What You'll Do

• Define the technical architecture and overall stack for an agent-native business applications platform spanning pricing guidance, booking and order intelligence, upsell recommendations, churn and retention prediction, media planning, deal scoring, revenue forecasting, and more.
• Evaluate LLMs, multimodal systems, multi-agent orchestration frameworks, and recommendation, ranking, and forecasting models for product use.
• Design and build the pipeline from business and customer signals to model inference, recommendation generation, output validation, and integration with internal revenue systems and APIs.
• Build production-grade systems with strong error handling, output validation, explainability, auditability, and human-in-the-loop guardrails for high-stakes pricing and financial decisions.
• Partner cross-functionally with ML, backend, frontend, data, and business teams to iterate quickly based on feedback and business needs.
• Drive technical decisions that directly influence revenue impact, product quality, scalability, and time-to-market.

Required Qualifications

• Bachelor's degree in Computer Science or a related field.
• Experience building recommendation or decisioning systems, ideally in advertising, media, or revenue-platform environments.
• Strong understanding of modern LLMs and agentic systems, and the ability to evaluate latency, cost, and quality tradeoffs.
• Solid experience with LLM and multi-agent pipelines, including prompting, tool use, orchestration, tradeoff analysis, and error handling.
• Experience deploying ML systems in production, including model serving, containerization, CI/CD, and monitoring.
• Hands-on experience with modern ML frameworks and tooling such as PyTorch, Hugging Face Transformers, agent orchestration frameworks (e.g., LangGraph or similar), feature stores, and vector databases for RAG workflows.
• Experience designing evaluation approaches for recommendation and generative systems using human review, automated and offline metrics, and online A/B testing.
• Strong software engineering fundamentals and solid production experience in Java or Python.
• Ability to translate ambiguous business requirements into practical technical solutions and communicate tradeoffs clearly to cross-functional partners.

Preferred Qualifications

• Startup or founding-engineer experience, or experience leading fast-moving Gen AI or agentic product initiatives.
• Experience owning a business-facing or self-serve product from architecture through production deployment and operational support.
• Deeper background in recommendation systems, pricing optimization, demand forecasting, uplift and churn modelling, or quoting and optimization.
• Experience improving inference cost and latency, and building responsible AI guardrails such as safety, compliance, auditability, and human review mechanisms.
• Open-source contributions or publications in ML, recommender systems, or applied AI for revenue and decisioning.