LLMOps Engineer

Описания нет

Select, benchmark, and deploy open-source LLMs and embedding models Serve models in production using vLLM, TGI, Ollama, or similar frameworks Run and optimize on-premises and private-cloud GPU infrastructure Tune latency, throughput, memory, and cost through quantization, batching, and caching Set up monitoring, logging, and scaling for model serving Work with AI and Dev Ops engineers to bring RAG and agent systems into production

Hands-on experience running LLMs in production on your own infrastructure, not only through OpenAI or Anthropic APIs Experience with vLLM, TGI, Ollama, Triton, or similar serving frameworks Solid Linux, Docker, and Kubernetes skills, including GPU workloads Python skills English at B2 level or higher Будет плюсом:
Quantization (GPTQ, AWQ, GGUF) and inference optimization, fine-tuning (LoRA / QLoRA), monitoring with Prometheus / Grafana, experience with air-gapped or highly secure environments

Freelance project Long-term collaboration opportunities Challenging AI projects to deepen expertise and develop technical skills Opportunities to work with modern AI technologies and solve real-world engineering problems A small, experienced team that values engineering quality