Senior AI Platform Engineer (LLM Serving & AI Operations)

Quick Facts

• Focus: LLM serving & AI operations in production
• Scope: AI runtime platform operations, observability, automation, reliability

Description

You will take a key role in operating an AI runtime platform—LLM serving, routing, vector databases, observability, and automation. This is a production-operations position for reliable model/service runtime, not pure research or prompt engineering. You will set operational standards and transfer know-how to the team developing AI managed services long-term.

Responsibilities

• Operate AI managed services: LLM serving, vector databases, monitoring, and routing
• Deploy and operate models and AI services in containers on Linux
• Develop automation tools in Python
• Build and maintain CI/CD pipelines and Infrastructure as Code
• Provide AI service observability: metrics, logs, dashboards, alerting, and routing tokens
• Ensure security and reliability of operations and protect sensitive data
• Handle incident management, capacity planning, and continuous cost/performance optimization
• Create runbooks, mentor teammates, and transfer knowledge
• Participate in on-call for AI managed services and related cloud components

Requirements

• Advanced Linux administration, automation, and Bash scripting
• Python experience for automation and work with AI/ML platforms
• Kubernetes and Docker experience in production or close to production
• Practical experience with LLM serving or AI inference workloads
• Knowledge of at least part of the stack: vLLM, Triton Inference Server, Ollama, or Hugging Face
• Observability and/or routing experience: Prometheus, Grafana, InfluxDB, LiteLLM, Langfuse, or Helicone
• Database fundamentals: PostgreSQL and/or vector databases
• Troubleshooting distributed services, incident management, and production support/on-call experience
• At least 3 years of relevant experience, independence, and mentoring ability

Benefits

• Work with modern technologies and large infrastructure
• Technical ownership of model serving and AI observability, with opportunity to take a modern LLM stack into commercial production operations
• Influence solution design from the start
• Small experienced team, minimal bureaucracy, space for initiative
• Hybrid work (home and office), flexible start/end times
• Indefinite-term contract and transparent bonus scheme
• 5 weeks of vacation + 5 MyDays; cafeteria, meal allowance, parking, and company recreation facilities
• Office location: Strahov (near Ladronka Park)