Glossary.
The words we use when we talk about AI interviews and tacit knowledge, in short. There are 26 of them, in four groups.
Updated on
The interview.
Tacit knowledge
What a person can do without being able to explain it. An adjustment made by feel, the client to call before the others, the exception no procedure covers. More in Tacit knowledge.
Explicit knowledge
Knowledge already written down in procedures, manuals and technical sheets. Anyone can read it and pass it on.
AI interviewer
An agent that runs the interview by voice or in writing. In Quisquilia it has a personality, a style, goals and quality indicators for the conversation. The Method explains how it works.
Elicitation
The techniques that bring out what a person knows but would not say on their own. Indirect questions, scenarios and follow-ups instead of blunt questions.
HUMINT
Human intelligence, the techniques of people who listen for a living. Part of our listening method comes from there.
Indirect question
A question that gets to the point through a concrete episode. It asks how it went last time instead of asking whether you are good at it.
Follow-up
The question that starts from the answer just given and asks for a detail, an example or the reason why. That is often where tacit knowledge comes out.
Scenario
A situation to imagine, offered instead of a direct question. It shows how a person reasons and decides.
What comes out.
Skills mapping
Hard and soft skills area by area, the skill gaps and a private profile for each person. Examples in Skills mapping.
Skill gap
The distance between the skills a role needs and the ones that come out of the interviews.
Training needs analysis
The study of what people need to learn to do their work well. It starts from the skill gaps and leads to the training plan. Consultants often run it with Quisquilia.
Job description
The description of a role with its activities, responsibilities and skills. With interviews it is written on how the role is really done. See Job descriptions.
Process mapping
The reconstruction of a process from the versions of everyone who works on it. Where the versions do not match you see it at once. See Process mapping.
AI readiness assessment
A measure of how much people use AI, how well they understand it and who can bring it to their colleagues. It comes before bringing AI into processes. See AI readiness assessment.
Rules and data.
Pseudonymisation
Replacing names and other data that identify a person before the analysis.
Aggregate result
A result about a group, not about a single person. Below the agreed threshold the group is not shown.
Individual result
A result about one person. Who receives it is declared before the project starts.
GDPR
The European regulation on personal data protection. How we apply it is in Compliance and security.
AI Act
The European regulation on artificial intelligence. For interviews what matters is who receives the results. If an individual result goes to a manager or to HR the system is high risk and we treat it that way.
DPIA
Data protection impact assessment. It is usually required when the processing is high risk.
Human checkpoint
The point where the analysis stops and a person checks before it goes on.
Working with Quisquilia.
Practitioner
Someone who brings listening to their own clients. Consultants, HR and training firms, research studios.
Powered by Quisquilia
The mark on partner projects. The client and the brand belong to the partner, Quisquilia runs under the hood. See For consultants.
Private label
Quisquilia without our mark. It is available to organisations that use it internally.
Probe
One topic, one question. A group of people is interviewed and you receive the structured conversations with a summary of the themes.
Continuous listening
The same listening repeated in waves over time. Each wave shows what changed since the previous one.