Workshop Overview

amberflo unified interface workshop this workshop walks you through configuring the amberflo unified interface from scratch and sending real ai traffic through it by the end, you will have a provider connected models exposed through the gateway a workload with secure access real api traffic flowing through amberflo usage and cost attributed by workload and model if you have questions, want to report a bug, or suggest features, join our amberflo discord community 👉 https //discord gg/femn96kv8z how this workshop is structured this workshop is broken into three focused sections 1\ get a working unified interface set up a provider, add models, create a workload, and send your first request from the playground 2\ integrate with code call the gateway from python or curl and verify usage data updates in real time 3\ track usage and cost across applications create multiple workloads and see how spend is attributed by workload and model each section builds on the previous one you can complete each independently prerequisites you will need an amberflo account at least one provider api key (for example, openai) basic familiarity with making api calls no prior experience with amberflo is required mental model amberflo separates configuration into four simple components provider grants access to an ai platform such as openai or aws bedrock model defines which llm is available applications reference models by alias workload represents an application or system that consumes ai workload key controls access and ties usage and cost to a workload this structure gives you unified access across providers consistent api structure clear cost attribution governance and control