Senior AI Inference Platform SRE & Automation

Bitdeer; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Bitdeer is seeking an experienced SRE to own the reliability of NeoCloud inference services. You will ensure end-to-end availability for model endpoints, cover the API gateway, router, and scheduling stack, and drive 99.95% reliability commitments.

The role involves multi-cluster deployments, GPU fleet health, and robust incident response in a fast-moving AI/Cloud environment. The ideal candidate has 5+ years in SRE, strong Kubernetes expertise, hands-on experience with LLM inference serving,

Qualifications

• 5+ years in SRE / production platform operations.
• 2+ years on customer-facing API services.
• Deep Kubernetes operations experience.
• Experience with LLM or model inference serving.
• GPU fleet operations and NVLink/NCCL fault isolation.
• Observability and incident response fundamentals.

Responsibilities

• Own end-to-end reliability of inference endpoints — API availability, latency, and SLA adherence.
• Manage the serving path: API gateway, inference router, scheduler, queue, and KV cache.
• Oversee model loading, readiness validation, and multi-cluster deployment for cross-region serving.
• Improve resilience with staged rollouts, canaries, and incident drills.
• Define and publish customer-facing SLOs/SLAs, and drive fault-tolerant improvements.

Skills

SRE

Kubernetes

Go

Python

Terraform

GitOps

LLM inference

APIs

GPU fleet

Tools

Prometheus

VictoriaMetrics

Grafana

DCGM

NCCL

Kubernetes

ArgoCD

Job description

Bitdeer is seeking an experienced SRE to own the reliability of NeoCloud inference services. You will ensure end-to-end availability for model endpoints, cover the API gateway, router, and scheduling stack, and drive 99.95% reliability commitments.

The role involves multi-cluster deployments, GPU fleet health, and robust incident response in a fast-moving AI/Cloud environment. The ideal candidate has 5+ years in SRE, strong Kubernetes expertise, hands-on experience with LLM inference serving,