— Glossary

Plain-English technology glossary.

Clear definitions of the terms we work with — Cloudflare, AI, infrastructure, and the words that come up when you are choosing a technology partner.

Cloudflare Workers: Cloudflare Workers is a serverless platform that runs your code on Cloudflare’s global edge network, close to users, rather than in a single regional data centre. Each request is handled by a lightweight V8 isolate that starts in well under a millisecond, so there are no cold starts and no servers to manage.
Edge computing: Edge computing runs application code in many locations close to where users are, instead of in one central server. For web applications it means lower latency and shorter round trips, because the response is generated near the visitor.
Production RAG pipeline Retrieval-Augmented Generation: Retrieval-augmented generation (RAG) is a technique where a language model answers using documents retrieved from your own data at query time, instead of relying only on what it learned in training. A production RAG pipeline is the engineered system around that — ingestion, chunking, embeddings, a vector store, retrieval, and guardrails — built to run reliably rather than as a demo.
MCP server Model Context Protocol: An MCP (Model Context Protocol) server is a small service that exposes tools, data, and actions to an AI assistant over a standard protocol, so the assistant can use them safely. It is how a model such as Claude reaches your systems — reading a database, calling an API, running a task — through a defined, auditable interface rather than ad-hoc glue.
Durable Objects: Durable Objects are a Cloudflare primitive that gives you a single, addressable piece of state with strong consistency, plus its own compute. They are how real-time and coordinated features — chat rooms, collaborative editing, rate limiters, game or session state — are built on Workers without a separate database for coordination.
Workers AI: Workers AI is Cloudflare’s platform for running machine-learning models — language models, embeddings, image and speech models — on its own GPU network, called from inside a Worker. It lets you add inference to an application without managing GPUs or routing every request to a third-party model provider.
Proxmox VE: Proxmox VE is an open-source virtualization platform for running virtual machines and Linux containers on your own hardware, managed from one web interface. It is a common, license-cost-free alternative to VMware for organisations that want to run their own servers.
Managed IT: Managed IT is an arrangement where an outside team takes ongoing responsibility for the technology a business runs on — devices, networks, security, backups, and support — for a predictable fee, instead of fixing things only when they break. The aim is fewer incidents and less downtime, not just a faster repair.
Technology partner: A technology partner is a senior team that takes end-to-end responsibility for designing, building, and operating the technology a business depends on — across web, software, infrastructure, and AI — rather than delivering one project and leaving. It differs from an agency, which is usually project- and marketing-led, and from a freelancer, a single person with a single skill, in both breadth and in staying responsible for what it ships.