When you interact with one of our AI agents, data is sent between a number of data processors.
ElevenLabs is used as the back-end system to provide the speech audio communication between you and the AI agent. It is responsible for converting audio from your voice into text, and text generated from the LLM (large language model) into synthesised voice audio.
The LLM, Anthropic Claude Opus 5, is the layer running below ElevenLabs. This takes raw text given by ElevenLabs’ speech-to-text, and uses a neural network to generate a textual response. This response is handed off to ElevenLabs to generate text-to-speech audio. The LLM can also trigger defined, programmed functions created by us (Robots of London) and agreed with the client within a client-side application or service, based on intents recognised in the LLM’s input.
We, Robots of London, do not store any audio or conversational data unless otherwise asked to.
