Senior Software Engineer at Playonsports in Remote

PlayOn Sports processes over 250,000 live high school sports games a year across NFHS Network, GoFan, and MaxPreps. We’re using AI and computer vision to turn those streams into real-time scores, player stats, automated highlights, and interactive fan experiences. As a Senior Engineer on the Streaming Intelligence team, you’ll build the human-in-the-loop and fan-in-the-loop interfaces that connect our computer vision pipeline to the people who use it — internal operators reviewing AI-generated stats and millions of fans engaging with AI-powered experiences across our three consumer brands. In this role, you’ll own the interactive layer between AI and users: annotation review tools, real-time stat overlays, correction workflows, and fan-facing features that run at scale across web and mobile. You’ll work primarily in Python, ship features end-to-end, and think AI-forward — not just consuming model outputs, but designing interfaces and services that make AI systems better through human feedback and fan interaction. The ideal candidate is a builder. You’re energized by shipping, comfortable with ambiguity, and excited about working at the intersection of AI and product. You don’t wait for a fully baked spec — you break down loosely defined problems and start delivering. THE OUTCOMES YOU’LL DELIVER • Production annotation review interface: Ship the first human-in-the-loop interface for the computer vision stats pipeline, enabling internal operators to review, correct, and approve AI-generated statistics in real time. Target: production-ready within six months. • Fan-facing AI feature: Deliver at least one AI-powered fan experience — real-time stat overlays, interactive highlights, or personalized content — to one of our consumer brands (NFHS Network, GoFan, or MaxPreps). Target: live within nine months. • Reusable AI development patterns: Establish the team’s standard UI component library and Python service templates for AI-forward development, enabling faster iteration on future human-in-the-loop and fan-in-the-loop features. • Correction workflow at scale: Build feedback loops that capture human corrections and fan interactions and route them back into the AI pipeline, measurably improving model accuracy over time. • Cross-brand consistency: Deliver interactive AI features that work reliably across NFHS Network, GoFan, and MaxPreps, adapting to each brand’s UX patterns while sharing a common service layer.