Observability Platform Engineer — Neocloud

About the Role Mirantis is building out our Neocloud service offering — managing large-scale infrastructure to a high SLA for customers running demanding compute workloads. As that offering scales, so does the volume and complexity of telemetry we need to collect, correlate, and act on. We're looking for an Observability Platform Engineer to design and build the monitoring, logging, tracing, and alerting platform that our operations teams depend on to detect and resolve incidents fast — at large scale, across a globally distributed environment.   This is a hands-on, build-it role. You'll be the person who turns "we have no visibility into this" into a platform that surfaces the right signal at the right time, and turns "we found out from the customer" into "we caught it before they noticed."   Key Responsibilities • Design, build, and operate observability platform components — metrics, logging, distributed tracing, and alerting — for large-scale infrastructure environments • Build telemetry pipelines capable of handling high cardinality, high volume data from large fleets of infrastructure, with an eye on cost, retention, and query performance • Define and implement SLO/SLI frameworks and alerting strategies that reduce noise and surface real signal to on-call engineers • Partner closely with service delivery and operations teams to understand what they need to see during an incident, and build for that — not just for dashboards nobody opens • Integrate observability tooling with incident management workflows, including root-cause analysis support and post-incident review data • Continuously improve detection speed and reduce mean-time-to-detect (MTTD) and mean-time-to-resolve (MTTR) across the platform • Contribute to the roadmap for AI-assisted operations tooling (e.g., automated triage, anomaly detection, engineer-assist tooling) as it matures • Own the reliability, scalability, and security of the observability stack itself — it needs to be up when everything else is on fire • Document architecture, runbooks, and operational practices so the platform is maintainable beyond you   What Success Looks Like • Operations teams can diagnose incidents faster because the right data is surfaced automatically, not hunted for manually • Alert volume is high-signal, low-noise — engineers trust what fires • The observability platform scales cleanly as infrastructure footprint grows, without cost or performance surprises • Reduced MTTD/MTTR trends, tracked and demonstrable over time • A platform other engineers actually want to build on, not work around • Proven experience designing and building observability platforms for large-scale, production infrastructure environments (not just consuming an existing setup — actually building one) • Strong hands-on experience with metrics, logging, and distributed tracing tooling (e.g., Prometheus, Grafana, OpenTelemetry, Loki, Thanos/Cortex/Mimir, Elasticsearch/OpenSearch, Jaeger/Tempo, or equivalents) • Experience with high-volume telemetry pipelines and the tradeoffs involved (cardinality, retention, cost, query latency) • Strong software engineering skills in at least one language commonly used in this space (e.g., Go, Python, Rust) • Experience with Kubernetes and cloud-native infrastructure • Solid understanding of SLO/SLI/error-budget practices and alerting design that minimizes noise • Comfortable working in a fast-moving environment where the platform is being built out alongside the infrastructure it's monitoring • Strong communication skills — able to work directly with operations/service delivery teams to understand real incident-response needs, not just technical specs Preferred Skills & Experience • Experience building observability for GPU/HPC infrastructure or o