Senior Software Engineer (W&B Agent)
Livingston • Permanent • Competitive
Back to Job Search

Senior Software Engineer (W&B Agent)

New Easy Apply
Livingston On-site Permanent 26 Applications
Competitive
Full-time
Posted 19 Nov 2026
Expires 19 Dec 2026

Job description

  • ARIA is the AI Research and Iterations Agent that enables continuous experimentation at scale. The engineering team is responsible for building this research agent from scratch by leveraging distributed backend systems that execute multi-turn agent conversations, to the React UI that thousands of ML practitioners interact with every day. Our mission is to ship a fast, reliable, and delightful agent experience that genuinely accelerates ML workflows, and every engineer on the team owns that outcome end to end
  • As a Senior Full Stack Engineer on the ARIA Agent team, you’ll own features and systems across the entire stack — from backend infrastructure and API design to the React frontend that powers the chat interface, billing UI, and agent configuration experience. You’ll build and maintain the distributed worker systems that run agent conversations at scale, design the streaming architecture that delivers real-time token output to the browser, and contribute to the frontend components that make ARIA feel fast and intuitive. You’ll collaborate closely with ML science, product, iOS, and SRE teams — blending reliability engineering with product sensibility. Day to day, you’ll move fluidly between debugging a production memory leak in a Kubernetes worker, shipping a new billing UI component, and hardening the egress controls on the code execution sandbox
  • Experience with observability tooling (Datadog, Weave tracing, structured logging) and a track record of building dashboards that surface actionable signals
  • You get energy from bridging theory and practice: designing scalable architectures and writing the code that proves them out
  • Strong Python backend skills — async patterns, FastAPI, distributed state machines, Kubernetes pod lifecycle
  • Background in ML infrastructure, developer tooling, or observability platforms
  • You’re curious about how systems behave under real-world load and enjoy building experiments that reveal what’s really happening under the hood
  • Proficient in TypeScript and React; comfortable building and maintaining production UI components including real-time streaming interfaces, complex state management, and responsive layouts across web and mobile web
  • You find satisfaction in long-term, cross-team collaboration — helping others succeed by embedding performance thinking into their workflows
  • Experience building streaming UIs (server-sent events, WebSockets) with graceful fallback handling
  • Security-conscious engineering mindset — input validation, sandboxed code execution, egress controls, and payload hardening
  • You’re an expert in transforming deep technical insights into tools and processes that empower other engineers to build faster, more reliable systems
  • Comfortable thriving in a fast-paced startup environment — able to move quickly, make pragmatic tradeoffs, and iterate on real user feedback without sacrificing production quality
  • Hands-on experience building and shipping production-grade AI agents — including tool use, multi-turn conversation management, and agentic reliability patterns
  • You love to turn messy, complex performance problems into clear, data-backed stories that lead to concrete action
  • Experience building and operating services at scale — rolling deploys, autoscaling, worker concurrency, graceful termination
  • Bachelor’s degree or foreign equivalent in Computer Science, Computer Engineering, Data Science, Information Systems, or a closely related technical or quantitative field involving substantial coursework in Engineering
  • Experience with MySQL or similar RDBMS for transactional state management
  • 4+ years of full stack engineering experience, with meaningful depth in both backend services and React/TypeScript frontend development
  • You care deeply about measurement, iteration, and continuous improvement — not just fixing bottlenecks, but preventing them from ever recurring
Is there something wrong with this job listing? Let us know.