blackforestlabs · 弗莱堡, 德国 · 2026 年 6 月 23 日(发布 61 天)
已经开放 61 天了。这类岗位通常在头两周就筛完人,现在投是逆风。
Member of Technical Staff - Research Infrastructure Engineer
这份工作
维护和扩展研究基础设施,包括集群、监控和性能优化,并参与轮班待命。
- Kubernetes
- Python
- Go
- Nvidia GPU drivers
- Prometheus
- SLURM
办公未说明
年限未说明
学历未说明
语言未说明
为什么是你
- Your current work at GMI Cloud — building an LLM proxy/gateway, tuning vLLM and SGLang, using OTel and dynamic routing — is almost exactly the 'research infrastructure' this team needs. You'd be scaling GPU clusters and optimizing inference systems, not starting from scratch.
- Python is the first language listed, and your entire hands-on history is in Python. The observability and Kubernetes side maps directly to what you've built with OpenTelemetry, Nacos, and Redis.
- This is a frontier AI lab where you could deepen your inference-serving expertise and stay in Germany. The role is infrastructure, not research, so your applied engineering background fits naturally.
我的顾虑
- The role is in Freiburg, not Munich, and hybrid on-site is required (2 days/week or a monthly in-person week). That means moving, which you'd rather avoid. The salary range starts at 100k, making it financially plausible, but the disruption is real.
- The posting emphasizes large-scale training platforms, while your LLM experience is on inference serving. The systems thinking transfers, but there may be a learning curve on the training side.
- Black Forest Labs is a private company, not a listed corporation. Your residence depends on the employer's stability; this is a prominent lab with commercial traction, but still a research startup environment.