Conversational
Responds to answers and asks meaningful follow-up questions, without becoming arbitrary.
thirdmind built an AI interview system for IRC Consult. Instead of asking for answers through a static form, it guides users through a real conversation.
Kolloqo is based on a structured questionnaire and keeps track of which information is needed to produce context-aware answers later.
A static questionnaire cannot follow up well. A human interview can, but is harder to scale. Generative AI can fill the middle ground: a structured conversation that works toward a defined goal.
Responds to answers and asks meaningful follow-up questions, without becoming arbitrary.
Keeps track of which information the questionnaire needs at the end.
No diagnosis, no therapeutic role. Professional responsibility remains with people.
Kolloqo guides the conversation, not a monologue.
Kolloqo is an AI interview system for structured surveys. It guides users through a conversation, checks completeness, and prepares answers for later evaluation.
Clear framework, clear goal of the survey.
Not everything at once, following the conversation.
Specific enough to close gaps without pushing users.
Sentences become questionnaire fields.
For later evaluation and domain classification.
The important point is that Kolloqo runs a clear conversation without losing the structure of the survey.
The list shows typical building blocks. Which steps are automated depends on the survey design and the desired analysis.
Sensitive topics need firm limits. A good interview system knows what it isn't.
Structural fidelity is not a detail.
An AI interview cannot be freely improvised. It needs a target catalog, follow-up logic, and clear limits for sensitive statements. Otherwise, the result is a pleasant chat, not usable interview data.
Defined fields that should be filled at the end of the conversation.
When to ask and when not.
How free answers become structured building blocks.
Clear boundaries instead of free interpretation.
Sensitive survey data remains in the EU.
Data does not flow into model training.
From survey design to productive dialogue.
We start with the survey logic: what information should be available in structured form at the end, in which context, and with which tone.
We then build a first version of the interview with defined follow-up logic and clear boundaries for sensitive topics.
If the structured answers arise reliably, the system is integrated into the survey process. Extensions follow iteratively.
thirdmind has built an AI interview system that guides users through a structured conversation and prepares questionnaire answers.
A questionnaire asks fixed questions. The AI interview reacts to answers, asks follow-up questions, and collects missing information in a targeted way.
Not across the board. It supports recurring survey processes. Professional responsibility and sensitive assessments remain with people.
No. Kolloqo is an AI system for structured interviews and questionnaire logic. It does not replace diagnosis, therapy, or professional assessment.
Because it can react to answers conversationally while still working toward a structured goal, as long as the interview logic is clearly designed.
No. Concrete figures, scientific comparisons, or effectiveness claims should only be mentioned after final approval.
Yuno asks four short questions about the survey objective, data, sensitivity, and evaluation. After that, it is clearer whether an AI interview makes sense.