JOBSEARCHER

Platform Reliability Engineer

Overview: In this role you will build and operate the observability backbone for an internal AgentCore platform on AWS Bedrock, making the platform visible and dependable. You will implement OpenTelemetry instrumentation, integrate with client observability tools, and define SLOs and cost monitoring to detect and fix slow, failing, or costly agent activities. You work closely with client engineers as part of a dedicated delivery pod, driving reliable, scalable observations and incident readiness. This is a client-facing, named-team engagement with a focus on practical, measurable reliability at scale.
Responsibilities:
  • Instrument OpenTelemetry instrumentation standards across agents, tools, and platform services and enable easy adoption by teams
  • Connect AgentCore Observability and CloudWatch to client observability tools (Datadog, Splunk, Grafana)
  • Implement end-to-end tracing across agent steps, model calls, tool calls, and multi-agent handoffs
  • Define SLOs for latency, availability, and error rates; build useful, low-noise alerting
  • Track token usage and compute spend by team/agent/environment with dashboards, budgets, and spend alerts
  • Prepare runbooks, support incident response, and lead post-incident reviews to drive real fixes
Key requirements:
  • 6+ years in SRE, platform reliability or observability engineering with strong hands-on AWS experience
  • Hands-on OpenTelemetry (SDKs, collectors, exporters) and distributed tracing
  • Experience with Amazon CloudWatch and at least one enterprise observability platform (Datadog, Splunk, Grafana, New Relic)
  • Experience defining and running SLOs, error budgets and alerting strategies
  • Experience building cost visibility and FinOps reporting on AWS (Cost Explorer, CUR, tagging strategies)
  • Scripting and coding skills in Python, Go or TypeScriptClear communication and ability to turn data into decisions for technical and non-technical audiences
  • Clear communication
  • Data-to-decision storytelling
  • Calm under incident pressure
  • OpenTelemetry (SDKs, collectors, exporters)AWS (S3, CloudWatch, Billing, IAM)
  • Cloud observability tools (Datadog, Splunk, Grafana, New Relic)