What is what.
The questions we get asked most, answered in two sentences before we explain the rest. Written for the person who has to decide, not for the person who builds.
AI baseline measurement
An AI baseline measurement records where an organisation stands with AI before anything is changed, so later progress can be compared against a fixed point. It covers current use, data quality, skills and decision-making, and its value lies entirely in being repeated later.
Read the full explanationAI Capability
AI Capability is an organisation's own ability to keep putting AI to work without outside help. It is not a tool you buy or a specialist you hire, but the combination of people, decisions, data and working methods that stays behind after the project is finished.
Read the full explanationAI diagnosis
An AI diagnosis is a structured examination of why AI is not delivering in an organisation, or of what would be needed before it could. It looks at process, data, decision-making and skills, and its output is a cause rather than a list of recommendations.
Read the full explanationAI literacy
AI literacy is the level of knowledge staff need to work with AI systems responsibly and judge their output. Since 2 February 2025 the EU AI Act obliges providers and deployers to ensure a sufficient level among everyone operating AI on their behalf.
Read the full explanationAI Photo
The AI Photo is Radical's free self-scan that shows an organisation where it stands with AI. It takes about ten minutes, asks no commitment, and produces a snapshot of your starting point: how decisions are made, what the data looks like, and which opportunities are within reach.
Read the full explanationAI policy
An AI policy is the internal document that records what an organisation allows with AI, who decides, and what staff may and may not put into a tool. It exists to make everyday use safe and reviewable, not to satisfy a regulator on paper.
Read the full explanationAI prioritisation matrix
An AI prioritisation matrix ranks possible AI applications against each other so an organisation can decide what to do first. It scores each opportunity on value and on feasibility, where feasibility covers the organisation as much as the technology: data, ownership and capacity to maintain.
Read the full explanationAI Readiness Scan
An AI Readiness Scan is a two-week diagnosis of how ready an organisation is to put AI to work. It maps where you stand, ranks the opportunities that matter, calculates the return on the strongest ones, and delivers a roadmap your board can actually decide on.
Read the full explanationAI roadmap
An AI roadmap sets out which AI applications an organisation will build, in what order, and what each one requires in people, data and decisions. A usable roadmap names an owner and a date per step, so it functions as a plan rather than a wish list.
Read the full explanationAI ROI
AI ROI is the return on an AI investment measured against its full cost, including licences, implementation, maintenance and the hours your own people spend. It is only meaningful when calculated on your own figures, because the same application returns very different amounts in two organisations.
Read the full explanationAPAC framework
APAC is Radical's own four-dimension assessment for judging AI professionals on qualities a CV cannot show. The letters stand for Adaptability, Personality, Awareness and Connection. It is deliberately named and structured, because vague criteria such as cultural fit are where bias hides.
Read the full explanationCapability Sprint
A Capability Sprint is the build phase in which Radical delivers the first working AI use case together with your own people. The point is not only the working application but the transfer: your team can run and change it afterwards without the external party.
Read the full explanationHuman on the loop
Human on the loop means an AI system runs on its own while a person supervises it and can intervene or overrule at any moment. It differs from human in the loop, where a person has to approve every individual action before it happens.
Read the full explanationSemantic layer
A semantic layer is a translation layer on top of existing systems that records what data means, so terms like customer, order or margin mean one thing everywhere. It lets AI work on your data without changing or replacing the underlying systems.
Read the full explanationTeam Effectiveness
Team Effectiveness is the optional final step of a capability trajectory, in which missing qualities are recruited so a new way of working is anchored durably in the team. It follows on only when the gap is a quality the team does not have, rather than a skill it can learn.
Read the full explanation