Applied AI infrastructure

Serving, scaling and operating models on GPU cloud infrastructure.

Advanced·6 weeks·8 modules·36 lessons

Module 1 is open

Start now with no account. One free account opens modules 2 to 8 here and in every other course.

Start module 1

The 8 modules

Each one is a set of lessons and ends with one thing you keep. Open a module to see its lessons.

01Architecture and workload discoveryWhat the workload actually needs, and what it costs and returns today.4 lessons · ends with your baseline and decision recordOpen
Write your baseline and decision record
02Compute, storage and networkingSizing the machine, and knowing the bill before it arrives.5 lessons · ends with your cost modelAccount
A free account opens this module and every other one.
03Packaging and reproducibilityA build somebody else can reproduce, including the prompts.3 lessons · ends with your versioned buildAccount
A free account opens this module and every other one.
04Serving and orchestrationThe serving path, from request to response, under load.5 lessons · ends with your serving stackAccount
A free account opens this module and every other one.
05LLM and multimodal deploymentThe choices specific to serving a language or multimodal model.3 lessons · ends with your model serving pathAccount
A free account opens this module and every other one.
06Release engineeringThe service goes to production behind a gate. Deploying it completes the course.5 lessons · ends with your release pipelineAccount
A free account opens this module and every other one.
07Observability and operationsWhat you watch, what wakes you, and what the numbers say against module 1.5 lessons · ends with your dashboard and runbookAccount
A free account opens this module and every other one.
08Security, governance and capstoneThe threat model, the evidence, and the record that closes the course.6 lessons · ends with your completion recordAccount
A free account opens this module and every other one.