Support Systems Architect

OpenAI OpenAI · AI Frontier · San Francisco, CA · User Operations

This role focuses on building and iterating on tooling, data flows, and processes to support a customer operations team at scale. It involves designing automated workflows, implementing LLM-powered automation, codifying incident detection and response, standing up data pipelines for insights, and identifying/solving risks during tooling rollouts. The goal is to redefine modern support organization through resilient systems and automation.

What you'd actually do

  1. Build “Day-1 enabled” workflows; role-tailored playbooks, content auto-diffs from source docs, and other workflows that have been taken for granted in typical Support organizations.
  2. Continuously automate repetitive touchpoints with scripts, agents, and LLM-powered flows; implement governance, observability, evaluation gates, and safe rollback.
  3. Codify detection (windowing, dedupe, thresholds), on-call handoffs, and post-incident learning loops that protect customer experience and SLAs.
  4. Prototype and learn quickly—leveraging ChatGPT, Jupyter notebooks, Retool, and other tools—to prove value before hardening with Engineering.
  5. Stand up data pipelines that capture sentiment, ticket trends, and BPO insights, routing actionable signals back to Product within hours—not weeks.

Skills

Required

  • 8+ years of experience in building tools for internal teams, especially within a customer support environment
  • Shipped or maintained tools and automations (dashboards, ETL pipelines, low-code apps) that eliminated manual work and scaled beyond a single team
  • Treat ChatGPT & LLMs as default co-developers, rapidly turning natural-language ideas into working code or queries
  • Deeply enjoy working cross-functionally and are skilled at building relationships with Product, Engineering, and Operations teams
  • Passionate about customer advocacy and have experience translating customer feedback into strategic product insights
  • Strong bias for automation and a distaste for doing low-complexity to otherwise repetitive work consistently
  • Thrive in a fast-moving, ambiguous environment where priorities will shift quickly and iterating on your systems will be required

What the JD emphasized

  • building resilient systems
  • automate repetitive touchpoints
  • LLM-powered flows
  • customer experience and SLAs
  • data pipelines that capture sentiment, ticket trends, and BPO insights
  • routing actionable signals back to Product

Other signals

  • building resilient systems
  • automate repetitive touchpoints
  • LLM-powered flows
  • customer experience and SLAs
  • data pipelines that capture sentiment, ticket trends, and BPO insights
  • routing actionable signals back to Product