Crusoe
Customer Success Manager, Managed Inference
$175K–$200K
San Francisco, CA - US · On-site · Manager · Full-time · Posted Aug 26, 2026
- Experience
- 2+ years; experience may be in customer success, technical account management, or a similar role supporting the listed customer types.
- Education
- Bachelor's degree in Business, Engineering, or a related field.
About the job
Supports customers deploying and scaling AI inference workloads on Crusoe’s cloud platform, providing technical guidance, relationship management, and coordination with internal teams.
What you'll do
- Build customer relationships and guide customers on cloud-based AI/ML solutions, including Kubernetes, model deployment, autoscaling, and observability.
- Review customer progress and performance indicators, including adoption, inference consumption, latency, and uptime; identify risks and growth opportunities.
- Coordinate customer feedback and issue resolution across internal teams, deliver training, and support escalations and service incidents.
What they ask for
- Bachelor's degree in Business, Engineering, or a related field.
- 2+ years of experience supporting enterprise cloud, AI, machine learning, developer platform, or infrastructure customers in customer success, technical account management, or a similar role.
- Technical foundation in cloud computing, AI, and ML; understanding of inference workloads, model serving architectures, Kubernetes, containers, APIs, and GPU infrastructure.
- Strong interpersonal, communication, and presentation skills; ability to communicate complex concepts simply.
Nice to have
- Familiarity with LLMs, RAG architectures, agentic applications, or inference optimization.
- Experience with Managed Inference, model serving platforms, GPU-based inference environments, or customer advocacy.
Pay
up to $175,000 – $200,000 OTE (total pay)
Tools
Kubernetes, LLMs, RAG
AI focus
Supports customers deploying, optimizing, and scaling AI inference workloads, including model serving, GPU utilization, and production readiness.
Model-written summary · the original posting is authoritative