Site Reliability Engineer

Cognition•San Francisco, CA

19d•$260,000 - $300,000

About The Position

We are an applied AI lab building end-to-end software agents. We're the makers of Devin, the first AI software engineer. Our team is extremely talent-dense. Among our founding team, we have world-class competitive programmers, former founders, and leaders from companies at the cutting edge of AI including Scale AI, Palantir, Cursor, Waymo, Tesla, Lunchclub, Modal, Google DeepMind, and Nuro. Building Devin is just the first step—our hardest challenges still lie ahead. If you’re excited to solve some of the world’s biggest problems and build AI that can reason on real-world tasks, apply to join us. Devin and Windsurf are used by hundreds of thousands of developers every day. When something goes wrong, it goes wrong for all of them at once. This role exists to make sure that doesn't happen, and when it does, to make sure it's resolved faster than anyone expects. You will own both the production reliability of our user-facing products and the platform engineering that lets our team ship quickly and confidently. That means SLOs, incident response, and on-call on one side, and CI/CD pipelines, deployment infrastructure, and developer tooling on the other. At Cognition, these are not separate jobs. The best SREs here understand that reliability is engineered in, not bolted on.

Requirements

Deep experience running production systems at scale: SLOs, error budgets, on-call rotations, and incident command
Strong software engineering fundamentals; SRE at Cognition means writing real code, not just configuring tools
Proficiency with cloud infrastructure (AWS, GCP, or Azure), container orchestration (Kubernetes), and infrastructure as code (Terraform or equivalent)
Experience building and owning CI/CD pipelines and deployment infrastructure for fast-moving product teams
Strong observability instincts: knows how to instrument systems, build useful dashboards, and design alerts that surface signal without generating noise
A track record of reducing toil systematically through automation, not just working around it
Comfort owning incidents end to end: detection, triage, mitigation, resolution, and postmortem
Enough product empathy to understand what reliability means from a user's perspective, not just an infrastructure one

Nice To Haves

Experience with developer-facing products or platforms is a strong plus

Responsibilities

Define and own SLOs, SLIs, and error budgets for Devin and Windsurf.
Build the monitoring, alerting, and observability systems that give the team a clear, honest picture of service health at all times.
Lead incident response with speed and clarity.
Run blameless postmortems that turn outages into durable improvements.
Build the runbooks and tooling that make on-call sustainable and effective.
Own the deployment pipelines, release infrastructure, and internal developer tooling that let the team ship fast without breaking things.
Reduce toil systematically so engineers spend time on work that matters.
Manage cloud infrastructure through code.
Build reproducible, auditable, version-controlled environments that scale with the product and eliminate configuration drift.
Model growth, forecast resource needs, and ensure the infrastructure stays ahead of demand.
Profile and improve system performance before users feel it.
Treat security not as a separate concern but as a reliability requirement.
Ensure that misconfigurations, vulnerabilities, and access failures are caught and remediated with the same urgency as outages.
Partner closely with product and engineering teams to build reliability in from the start.
Be the person who catches the single point of failure in the architecture review before it becomes a page at 2am.