AI at Zapier
At Zapier, we build and use automation every day to make work more efficient, creative, and human. So if you’re using AI tools while applying here - that’s great! We just ask that you use them responsibly and transparently.
Check out our guidance on How to Collaborate with AI During Zapier’s Hiring Process, including how to use AI tools like ChatGPT, Claude, Gemini, or others during our hiring process - and when not to.
Job Posted: July 10th, 2026
Location: NAMER
Hi there!
As Zapier expands into the enterprise market and accelerates AI-driven development, incident management is a front-line function for customer trust and operational reliability. The Incident Operations Specialist role is the engine that keeps the day-to-day operation of Zapier's incident program running — reliably, visibly, and at scale.
You'll report to the Incident Program Manager and operate under their direction, owning the execution of the systems, workflows, and tooling that power incident response across Zapier. You are the person in the operations center: you keep things moving, you surface what's broken, and you use AI to build automation that lets a small team operate like a much larger one.
This isn't a pure engineering role; it's an ops role with technical depth. The right person thinks in systems, writes automation to scale their own work. They use AI as a default, build repeatable workflows, and leave things more reliable than they found them.
The role is particularly exciting because decisions here ripple across Zapier, shaping how we manage a key process that involves the whole company.
About You
- You run operations with precision. You keep complex systems running. You understand how incident tooling (incident.io, PagerDuty, observability, integrations, Slack workflows) plugs together, you notice when something breaks before anyone else does, and you fix it. Configuration, routing, escalation paths, on-call schedules — this is your domain.
- You build with AI, not just prompt it. You use AI-native tools (Cursor, Claude, Copilot, or similar) as your standard working environment, not as a novelty. You've built repeatable AI-powered workflows — things that keep running when you're offline. You know when an AI output needs verification, and you've built quality checks into your systems. You can quantify how your AI usage has actually changed throughput or quality.
- You're technical enough to operate the tools and build custom ones. You're not a software engineer, but you can build custom automations, write SQL to pull from Databricks, configure API integrations, and prototype lightweight AI agents to extend your own capabilities. You operate comfortably in GitLab, Coda, Slack APIs, and observability tools. If a workflow doesn't exist, you build it. If a dashboard is broken, you fix it.
- You close the loop. You finish what you start without needing follow-up. Jira reflects reality. Commitments land on time. When something slips, you flag it early; you don't go quiet.
- You prioritize ruthlessly. You receive requests from multiple directions. You apply judgment, push back on low-priority work that doesn't align with program goals, and protect your capacity for high-impact operational work. You clearly and quickly escalate trade-offs rather than getting pulled thin.
- You understand that incident rotations mean incidents are shared responsibility. You're empathetic with commanders under pressure, give feedback that improves future response without creating friction, and embrace feedback on your own work. You make the whole community better — you don't position yourself as the single point of expertise.
- You distill complexity into clarity. You're curious and resourceful. You ask the right questions to understand customer and technical impact, then translate that into plain-language guidance for Support, GTM, and leadership — without losing the signal.
- You work async-first. Zapier is 100% remote. You write clearly and proactively. You design your work to be transparent and handoff-ready. You know when to escalate to a live conversation and when to make the call and document it.
Things You'll Do
- Own incident tooling operations. Maintain the reliability and configuration of incident.io, PagerDuty, Slack-based workflows, on-call rotations, and escalation paths. Monitor integrations and automations for issues. Fix what breaks.
- Build and maintain AI-powered workflows. Design and ship repeatable automation: incident thread summarization, postmortem draft generation, follow-up triage, severity classification, and data hygiene workflows. Turn one-off experiments into durable systems that compound over time. Keep improving them.
- Build and sustain the IC community. Grow and maintain a community of practice for Incident Commanders and Support Leads. Run regular touchpoints, share learnings across incidents, and coach responders on what good looks like. You are a resource others come to — not just a system operator.
- Operate data and reporting systems. Build, maintain, and troubleshoot dashboards and reports (Databricks, Grafana, Looker). Ensure data quality, field completeness, and metric accuracy. Surface trends and operational signals to the Incident Program Manager before they become problems.
- Maintain documentation and enablement assets. Keep playbooks, templates, and incident guides current and usable under pressure. Flag gaps where program-level guidance needs updating. Be the go-to resource for questions about tooling and process.
- Drive continuous improvement. Participate in incidents and postmortem reviews. Identify patterns across incidents. Implement improvements based on hands-on observation. Surface recurring friction to the Incident Program Manager with recommended solutions, not just problems.
- Partner across the org. Incidents touch Engineering, Support, GTM, Legal, and Finance. You coordinate across these teams without needing the Program Manager to broker every conversation. You know who to loop in, when, and how to communicate what they need to act.
- Support the incident program at scale. As the program expands to cover Support, Legal, PR, and Finance, help onboard and enable new stakeholder groups. Build the infrastructure that lets the program run self-sufficiently — including during your own PTO.
Our Stack & Tools
- Incident tooling: incident.io, PagerDuty, Slack
- Data & Reporting: Databricks, Grafana, Looker, SQL
- Observability context: Datadog, Grafana, Prometheus, Opensearch, Graylog
- Collaboration: GitLab, Coda, Google Workspace, Jira, Zendesk
- AI tooling: Cursor, Zapier AI, Claude, or equivalent
Experience and Requirements
Technical and operational experience
- Experience in incident response, technical operations, or a reliability-adjacent role
- Hands-on familiarity with incident tooling (incident.io, PagerDuty or equivalent), and Slack-based workflow automation
- Comfortable building with SQL and reporting tools (Databricks, Looker, Grafana)
- Able to diagnose operational issues using logs, system context, and integration debugging
- Can build lightweight automations, configure APIs, and prototype AI workflows — doesn't require an engineer to do it for them
AI fluency (required, not optional)
- Demonstrably uses AI in daily work: for drafting, summarizing, building workflows, and triaging
- Has built repeatable AI-powered workflows (not just one-off prompts)
- Applies verification and judgment to AI outputs — especially in high-pressure incident contexts
- Can articulate how their AI usage has improved speed, quality, or operational capacity
Operations and execution
- Independently drives problems to resolution when direction is clear but the path requires investigation
- Strong attention to detail in configuration, data quality, and documentation
- Works visibly — status in public channels, proactive updates, no need to chase
- Proven track record of respectfully disagreeing and committing — you raise concerns early, then execute once a decision is made.
Communication and collaboration
- Clear written communication across orgs like engineering, support, and ops audiences
- Proactive in surfacing risks, blockers, and improvement opportunities
- Comfortable pushing back on off-program requests and escalating trade-offs
- Able to communicate highly technical incident information to non-technical audiences — Support teams, GTM partners, and leadership — clearly and without jargon.
What Success Looks Like
- Incident tooling runs reliably with no surprises. Routing, escalation, and on-call systems behave as expected.
- Dashboards and operational reports are accurate, up to date, and trusted by the Incident Program Manager and leadership.
- AI-powered workflows measurably reduce manual effort without degrading the quality of what they touch. Automation that saves time at the cost of accuracy or judgment isn't a win.
- Work is visible in public channels without prompting. Stakeholders always know what's in motion.
- Documentation and playbooks are maintained and usable under pressure.
- The program can operate independently during PTO and through role transitions; you've built systems that don't depend on any individual to keep running. Work is documented, config is explained, and runbooks cover the program itself, not just incidents.
Application Deadline:
The anticipated application window is 30 days from the date job is posted, unless the number of applicants requires it to close sooner or later, or if the position is filled.
Even though we’re an all-remote company, we still need to be thoughtful about where we have Zapiens working. Check out this resource for a list of countries where we currently cannot have Zapiens permanently working.
