Senior Machine Learning Engineer at Capital Group Companies in Seattle, WA

Job description

“I can be myself at work.” You are more than a job title. We want you to feel comfortable doing great work and bringing your best, authentic self to everything you do. We value your talents, traditions, and uniqueness—and we’re committed to fostering a strong sense of belonging in a respectful workplace. We intentionally seek diverse perspectives, experiences, and backgrounds, investing in a culture designed to celebrate differences. We believe that belonging leads to better outcomes and a stronger community of associates united by our mission. At Capital, we live our core values every day: Integrity, Client Focus, Diverse Perspectives, Long-Term Thinking, and Community. “I can influence my income.” You want to feel recognized at work. Your performance will be reviewed annually, and your compensation will be designed to motivate and reward the value that you provide. You’ll receive a competitive salary, bonuses and benefits. Your company-funded retirement contribution will factor in salary and variable pay, including bonuses. “I can lead a full life.” You bring unique goals and interests to your job and your life. Whether you’re raising a family, you’re passionate about where you volunteer, or you want to explore different career paths, we’ll give you the resources that can set you up for success. Enjoy generous time-away and health benefits from day one, with the opportunity for flexible work options Receive 2-for-1 matching gifts for your charitable contributions and the opportunity to secure annual grants for the organizations you love Access on-demand professional development resources that allow you to hone existing skills and learn new ones “I can succeed as a Machine Learning Engineer at Capital Group” We are looking for someone who can take a vague question from an investment professional, find a real answer in messy data, prove the answer holds, and build the thing that delivers it. You will join the AI Insights team. We build the insight layer on top of Capital Group’s investment data: multi-agent systems that answer investment questions with citations, the evaluation methods that tell us whether those answers are any good, agents that take on expert analyst workflows end-to-end, and the extraction work that turns unstructured research, calls, and filings into reusable insight. Some problems here are better served by a conventional supervised model, and part of the job is knowing whic h is which. Our work reaches across the investment organization, from research analysts to governance specialists to the teams behind portfolio and order management. Each partner brings its own data, its own workflow, and its own idea of what a good answer looks like. You go deep with one rather than skim, and you end up learning parts of the business most engineers never see. Whether systems like these actually work, and how anyone would know, is still an open problem in this field, and making it answerable here is a large part of this role. This is applied science with a delivery bar, not a research lab: the answers have to hold up to people making real investment decisions, and they have to arrive as something working rather than a paper. Everyone on this team builds. There is no version of this role where you hand a design to someone else and review what comes back. We work h and in hand with a partner engineering team that owns the platform, so your time goes to the insight and the evaluation rather than the infrastructure underneath it. “I am the person Capital Group is looking for.” You will: Sharpen an underspecified ask into a problem worth solving: what is really being asked, what would count as an answer, what evidence would settle it. Pull signal out of messy, incomplete data, and tell a real result from leakage, a lucky split, or a metric that flatters itself. Design the evaluations that tell us whether a Generative AI system is working: eval sets, success criteria, LLM-as-judge and its failure modes, and the judgment to kn