← Casesmentoring
20+ one-on-one consultations
Consultations on ML infrastructure through getMentor and requests from the channel. Engineers come to move into MLOps, and teams come to build ML infrastructure.
Numbers
What people come with
- "Where do I grow, and what does an MLOps engineer actually do?" The market itself is not sure: job titles range from DevOps to Technical Product Manager.
- How to build a platform for ML workloads and what to do with GPUs in Kubernetes.
- How to size infrastructure for LLMs and deploy inference on GPUs.
What they leave with
- A clear specialization: MLOps split into Data, ML and Inference, with one of them picked.
- A personal roadmap: what you already know, where the gaps are, what to learn and in which order.
- Materials after the session: articles, talks, checklists.
What we do in a session
We split MLOps into Data, ML and Inference. We go through the verticals and the horizontal layer that automates delivering value from data to the endpoint, and pick a specialization that fits the person's goals.
We build a personal roadmap: what you already know, where the gaps are, what to learn and in which order.
We dig into engineering questions: ML platform architecture, GPUs in Kubernetes, choosing GPUs and configurations for LLMs, inference services.
I share materials after the session: articles, talks, checklists. The best answers become public guides, like the one on deploying inference on a GPU VM.
Testimonials
Anton does not just know the subject - he is a practicing professional, and his answers come from hands-on experience.
George
Helped me sort out what I already know, what I need to learn and in which order.
Konstantin
Great understanding of GPUs, and he can explain everything clearly.
Vlad
Translated from Russian.
Topics
MLOps roadmapML platformsGPUs in Kubernetesinferencesizing infrastructure for LLMs
My role
Mentor on getMentor, consulting since February 2025.