Forward Deployed Engineer - AI Inference

The role

You will assist enterprises in deploying, scaling, and operating generative and agentic AI workloads on FriendliAI infrastructure. You will work directly with customers to solve and implement production-grade applications.

What you will do

• Design and implement large-scale deployment architectures for LLM and multimodal inference
• Deploy and manage containerized workloads across Kubernetes clusters
• Diagnose production issues and implement temporary fixes
• Collaborate with customers' DevOps teams to integrate infrastructure into CI/CD workflows
• Develop scripts, Helm charts, and Terraform modules
• Contribute field insights to shape platform reliability, observability, and scaling strategies
• Lead workshops, technical sessions, or webinars

Required qualifications

• 3+ years of experience in cloud infrastructure, DevOps, or reliability engineering
• Bachelor's or Master's degree in Computer Science, Computer Engineering, Electrical Engineering, or equivalent
• Proficiency with Kubernetes, Docker, Terraform, and Helm
• Strong foundation in distributed systems, networking, and performance tuning
• Experience with GPU-based computing and generative AI model serving workloads
• Strong technical background in backend systems or AI tooling
• Experience operating workloads on AWS, GCP, or OCI
• Excellent problem-solving and debugging skills in real-world environments

Nice to have

• Experience deploying large models (LLMs, diffusion models) on GPUs or clusters
• Familiarity with inference frameworks (Triton, vLLM, TensorRT, DeepSpeed-Inference)
• Familiarity with observability stacks (Prometheus, Grafana, Loki, ELK, OTEL)
• Understanding of networking security and compliance frameworks (e.g., SOC 2)
• Experience supporting on-prem or hybrid-cloud deployments

Benefits

• Competitive compensation and benefits package
• Daily lunch and dinner provided
• Unlimited snacks and beverages
• Health check-up
• Top-tier hardware support
• Flexible working hours

About the company

FriendliAI is the fastest inference cloud for agents, built to run frontier open-weight models in production at scale. It is a small, fast-moving team building a platform for generative AI infrastructure.