Senior AI Telemetry & Observability Engineer

Bitdeer (NASDAQ: BTDR); Singapore; On-site; SGD 120,000 - 240,000; Full time

Job summary

Bitdeer, headquartered in Singapore, is seeking an experienced observability engineer to architect and scale a high-cardinality telemetry stack for AI and GPU workloads. You will integrate NVIDIA DCGM and other exporters into Kubernetes, build eBPF-based diagnostics, and drive real-time analytics and billing metrics.

You should have extensive Prometheus/OpenTelemetry experience, strong Go and Kubernetes skills, and a track record of scaling telemetry in HPC or cloud environments.

Qualifications

• Requires strong experience in Prometheus/OpenTelemetry ecosystems.
• Proficiency in Go and custom Kubernetes metric exporters is essential.
• Hands-on with kernel tracing tools (eBPF/BCC) and Linux performance tuning.

Responsibilities

• Architect and scale telemetry infrastructure for high ingestion rates.
• Integrate hardware exporters into Kubernetes observability stack.
• Create eBPF-based tools to trace network and I/O latency across clusters.
• Develop dashboards and alerting pipelines for proactive maintenance.
• Design multi-tenant real-time GPU/network utilization billing metrics.
• Collaborate with GPU Systems teams to define observability standards for AI workloads.
• Lead design reviews and mentor on high-performance telemetry collection.

Skills

Prometheus

OpenTelemetry

Go

Kubernetes

eBPF

Kernel tracing

GPU telemetry

Education

Bachelor's or Master's in CS/EE

Tools

NVIDIA DCGM

Redfish

VictoriaMetrics

Thanos

Mimir

Job description

Bitdeer, headquartered in Singapore, is seeking an experienced observability engineer to architect and scale a high-cardinality telemetry stack for AI and GPU workloads. You will integrate NVIDIA DCGM and other exporters into Kubernetes, build eBPF-based diagnostics, and drive real-time analytics and billing metrics.

You should have extensive Prometheus/OpenTelemetry experience, strong Go and Kubernetes skills, and a track record of scaling telemetry in HPC or cloud environments.