Senior Site Reliability Engineer

Xplor Technologies powers the experiences at the heart of everyday life. Through modern vertical software, embedded payments, and AI-powered capabilities, we help businesses in fitness, recreation, golf and club, field services, laundry, education, and other membership-based and service-based industries simplify operations, uncover insights, and elevate customer and member experiences.

AI Engineering builds the shared agent platform behind Xplor's AI capabilities: the agent runtime, model gateway, retrieval, guardrails, evaluations, and cost controls. We build the hard parts once, so every product team can ship AI features faster and safely. We're looking for a hands-on Site Reliability Engineer to own how our agent platform behaves in production.

As AI agents take on more of our operational work, you'll decide what they're allowed to do and make sure they do it safely. You'll treat AI workloads as first-class engineering problems, including long streaming responses, background agents, model-provider quotas, outputs that can fail silently, agents that take real actions, and spend that grows with usage.

Responsibilities:

• Define SLOs and error budgets for agent APIs, streaming and background work, including quality signals like task success, tool-call failures and evaluation regressions, and alert on what matters to customers.
• Own the platform's AWS infrastructure (ECS on Fargate, Aurora PostgreSQL, Redis and Amazon Bedrock) as Terraform delivered through Spacelift.
• Build observability in Datadog and Coralogix with OpenTelemetry, including AI signals like token usage, provider latency and cost per task.
• Make data stores resilient, with tested backups and restores, failover, defined disaster recovery targets (RTO/RPO) and safe migrations.
• Build resilience into the model layer: provider failover, quota and rate-limit management, and safe rollout and rollback of model and prompt changes.
• Protect the platform under multi-tenant load with per-tenant quotas, isolation and circuit breakers that stop runaway agents and runaway spend.
• Partner with Security on operational controls for AI risks, including prompt injection, excessive agent permissions and unbounded consumption (OWASP Top 10 for LLM Applications).
• Set the guardrails for AI agents that work on our infrastructure: what they can change, which actions need human approval, and how every action is audited.
• Work with Security and Privacy to keep model inference and customer data in the right regions for customers in Australia, New Zealand, the US, the UK and the EU.
• Lead incident response and blameless reviews, take part in on-call, and run game days for the failures that matter most.
• Make deployments safe by default, and give product teams simple, standard ways to run agent workloads.
• Track and reduce cloud and model spend.

Our stack:

• Cloud: AWS: ECS on Fargate, Aurora PostgreSQL with pgvector, ElastiCache Redis, CloudFront, Route 53, Secrets Manager and SSM
• AI: Amazon Bedrock and other model providers behind a multi-provider gateway
• Infrastructure as code: Terraform, delivered through Spacelift
• Observability: Datadog, Coralogix and OpenTelemetry
• Delivery: GitHub Actions, containers, Amazon ECR and preview environments