process variables
two years of process, in a single image.
A therapist knows when their consultant is changing. They hear it in the voice, in what they bring, in how they tell their story. But that knowledge lives in memory, and memory compresses two years into an impression. Process variables do something else: they read each complete session and score five dimensions of the therapeutic work, with evidence cited to the minute. Session by session, the series draws the shape of the process — a shape you can look at, discuss and teach.
one real case, twenty-eight sessions, two years.
This is not a mock-up. It is the process of Maribel, a real educational case — octubre 2001 — octubre 2003 — shared with consent more than two decades ago for therapist education, and reanalysed with nexmin in 2026. The coloured line is the process index: the weighted average of the five variables. It starts at 37 and ends at 79.
Each ring marks a session where a life event surfaced. Only some carry a label: the detail of a biography does not belong on a public page, even with consent and twenty years of distance.
how to read the curve, and why it is not straight.
No real process rises in a straight line. This curve falls when the father dies. It falls in the hard cluster of the second autumn — sick leave, a hospitalised husband. From the second year onwards it does something no memory impression can prove: it stabilises high while life keeps hitting. The person loses their job in session 22 and the index barely moves. That — the difference between being well and being able to face what comes — is what a therapist wants to see, and it is almost impossible to see without the complete series.
One important nuance: the variables do not measure how life is going for the person. They measure how they work in session — who leads the story, what they do with what they feel, how much they bring. That is why the curve can stay high in objectively hard weeks: what has become solid is the capacity to process, not the circumstances.
the five variables, one by one.
agency
30 → 80 in this casefrom helplessness to authorship
Who leads the story. Whether the person tells it as someone things happen to or as someone who acts with what happens.
flexibility
30 → 75 in this casefrom rigidity to flexibility
The ability to look at the same thing from another place: nuance a belief, try another reading, hold two truths at once.
emotional processing
35 → 75 in this casefrom overwhelm to integration
What the person does with what they feel in session: whether emotion carries them, they avoid it, or they can hold it in front of them and work with it.
alliance
50 → 85 in this casefrom distance to mature alliance
The quality of the working relationship as it sounds in conversation: agreement on the task, trust in the therapist, capacity to disagree.
active participation
50 → 85 in this casefrom low involvement to protagonism
How much the person brings to the session: initiative, between-session work, their own questions.
what these figures are not. deliberately.
They are not a psychometric test or a clinical measurement instrument: they are computational estimates to guide your listening, with each score anchored to evidence you can verify at the exact minute of the audio. They do not diagnose. They do not predict relapse or discharge. They do not replace your judgement: the AI proposes, you decide. The curve is not the process; it is a map of the process, and the territory remains the person in front of you.
frequently asked questions.
How are these scores calculated?
The AI reads the complete session transcript and scores each variable from 0 to 100, citing evidence with its minute so the therapist can check it against the real audio. The process index is the weighted average of the five. Everything goes through the therapist’s review: the AI proposes, you decide.
Is it a validated psychometric instrument?
No, and we say so plainly. These are computational estimates meant to guide the therapist’s listening between sessions — a clinical compass, not a test. They do not diagnose, predict or replace the judgement of the person who holds the relationship.
Where does the chart data come from?
From a real educational case: 28 sessions between 2001 and 2003, shared with consent more than two decades ago for therapist education, and reanalysed with nexmin in 2026. The data is frozen on this page; no database or active case is queried.
Does it work with only a few sessions?
Session-by-session reading works from the first one. The curve — the shape of the process over time — gains meaning after a few sessions: the series is where valleys, recoveries and what holds become visible.
see the shape your next process is taking.
14 days free, no card