In this article
Model serving platforms handle model loading, request batching, GPU memory management, auto-scaling, and health monitoring. Options range from managed APIs (OpenAI, Azure OpenAI, Bedrock) to self-hosted solutions (vLLM, TGI, Triton, Ollama). Key decisions include latency requirements, cost targets, data residency, and whether to use managed vs. self-hosted inference.
What it means in practice
Model Serving is not just vocabulary; it is a design handle. Use it as a reference point when comparing architecture choices, debugging implementation trade-offs, or explaining system behaviour to another engineer. It helps teams choose the right model/runtime balance across quality, speed, memory, governance, and cost.
Why engineers care
- It gives teams a shared name for the behaviour, risk, or architecture choice being discussed.
- It helps separate the goal from the implementation detail, so you can compare alternatives instead of copying a tool pattern blindly.
- It creates a useful checklist for reviews: inputs, outputs, failure modes, ownership, cost, latency, and measurement.
Production watch-outs
Benchmarks are only a starting point. Validate with your prompts, data, latency budget, concurrency pattern, and safety requirements.
Related context
Useful neighbouring concepts: Inference, Quantisation, Vllm.