AI agent
A system that understands context, uses knowledge and can trigger actions instead of only generating text.
The glossary is designed for decision-makers, teams and clients who need a shared language without wading through unnecessary jargon.
A surprising amount of friction comes from terms that different stakeholders interpret differently. The glossary creates a clean reference layer inside the same brand language as the rest of the site.
Short, precise definitions written so technical and non-technical stakeholders can still make sound decisions from them.
A system that understands context, uses knowledge and can trigger actions instead of only generating text.
Retrieval-augmented generation combines a language model with a controlled knowledge layer and answers from actual sources.
A large language model such as Llama or Mistral that processes and generates language and powers many modern agent systems.
Software or models run on your own or deliberately managed infrastructure instead of public consumer services.
The ability to control hosting, permissions, data paths and access both technically and legally.
An open-source workflow automation platform that can be self-hosted and extended in privacy-conscious environments.
An AI system that can handle phone conversations, qualify callers and hand over cleanly to humans when needed.
The explicit escalation point at which an agent transfers the conversation to a human together with full context.
The structured assessment of whether an enquiry is operationally, commercially and strategically aligned with an offer.
The glossary will later expand into audit, visibility and process terminology so website, sales and delivery all speak the same language.
If the bottleneck is still unclear after that, audit is usually the fastest way to prioritise the next move properly.