MI Practice Lab

Keep MI skills alive between workshops

A realistic AI client, structured MI feedback, and a "share with coach" handoff that turns every session into a teachable moment. Built to extend the workshop, not replace it.

MI Practice Lab session screen
The problem we're solving

MI skills decay without practice

Decades of research on MI training tells the same story: a two- or three-day workshop produces real gains in MI competency, and those gains erode within months unless practitioners practise.

Miller & Mount (2001) found that workshop-only training produced statistically significant skill gains that did not translate into changes in client outcomes — practitioners said the right things less often than they thought, and stopped saying them shortly after the workshop ended.

Multiple follow-up studies have replicated the finding: without ongoing practice and feedback, MI skills decline. The workshop is the on-ramp. What happens next determines whether the skill survives.

What the research consistently shows

  • Workshops alone are insufficient for sustained MI proficiency.
  • Coaching and structured feedback are the strongest predictors of skill retention.
  • Self-assessed MI skill systematically overestimates measured MI skill.
  • Repeated practice with feedback is what closes the gap.

Miller & Mount (2001); Madson, Loignon & Lane (2009); Schwalbe, Oh & Zweben (2014).

What learners actually do in the Lab

Six things that make the Practice Lab a serious training tool, not a chatbot wrapper.

Scenario picker showing clients including Adolescent Suffering from Depression and Adolescent Vape Dependence
01

A library of clinically realistic scenarios

Pre-built clients with real clinical concerns — adolescent substance use, depression, smoking cessation, weight management — each with their own ambivalence, history, and willingness to engage. Trainers can also commission bespoke scenarios.

Live coaching panel showing real-time feedback during an MI session
02

Live coaching while you practise

Toggle Guided or Deep-dive coaching mid-session. As you speak, the Lab flags closed questions, missed change talk, and missed opportunities for reflection — and suggests what to try instead. Or turn it off and run unassisted.

Session review showing MI Score 62, MI Spirit pillars rated 4/5 and 3/5, and Talk Ratio at 48.4% / 51.6%
03

MI Score and MI Spirit, broken down

Every session ends with an MI Score, a balanced Talk Ratio reading, and an MI Spirit breakdown across the four pillars: Partnership, Acceptance, Compassion, Empowerment. Each with its own rubric and per-pillar feedback.

Session playback timeline with OARS markers and Engage Focus Evoke Plan structure
04

OARS-tagged session playback

Replay the conversation with every Open Question, Affirmation, Reflection and Summary marked on the timeline. The Engage · Focus · Evoke · Plan structure is overlaid so you can see where the session moved — and where it stalled.

What went well and Areas for improvement tabs showing concrete examples with timestamps
05

What went well, where to grow

Specific, evidence-anchored feedback citing exact moments in your transcript. Skilful reframes you nailed, sustain talk you missed, and partial commitments that needed pinning down — with timestamps you can jump to.

Critical moments list including Excellent MI, Strongest Moment, and Practice Spot tags
06

Critical moments worth revisiting

The Lab surfaces five-to-ten pivotal moments per session — Excellent MI, Strongest Moment, Practice Spots — so you don't have to re-listen to the whole conversation to find what mattered.

For trainer-led cohorts

Share the session with a coach

One tap sends the full transcript and audio recording of a session to a designated coach or trainer. The coach receives an interface to leave detailed feedback inline — anchored to specific moments in the conversation.

That turns the Lab into a structured between-session check-in tool: trainees practise, share, get nuanced human feedback, and bring richer questions to the next workshop session.

See how trainers use this

Full transcript

Time-stamped, OARS-tagged, searchable.

Audio recording

Listen back to tone, pacing, and moments of empathy.

Inline coach feedback

Coaches respond against specific moments — not the whole session in one go.

Evidence

Is the AI's MI scoring any good?

Early bench testing shows promising alignment between the Lab's MI classification and human MITI-trained coders on core OARS behaviours. The numbers aren't yet at publication-grade — that's deliberate; we'd rather ship caution than overclaim.

A more rigorous evaluation against MITI 4.2.1 is in progress, and we'll publish the methodology and results in full when they're ready. In the meantime, treat the Lab's scores as formative practice feedback — not a credentialing assessment.

What we're testing

  • · Agreement with MITI-coded reference sessions
  • · OARS classification accuracy (open vs closed questions, simple vs complex reflections)
  • · Change-talk vs sustain-talk discrimination
  • · Stability across client demographics and presenting concerns

If you're an MI researcher and want to see the methodology in progress, get in touch.

Coming soon

CBT scenarios — same Lab, same coach feedback loop

We're extending the Practice Lab into Cognitive Behavioural Therapy: same interface, same share-with-coach handoff, same session structure adapted for CBT. If you train CBT and want early access, come and talk.

Already in trainers' cohorts

Her son walked in, heard the conversation, and quietly walked back out. He thought she was talking to a real patient.
— MI trainer using the Lab with her cohort
This is the future of MI training.
— MI trainer running cohort-based programmes

Two of the trainers running early cohort pilots. Named testimonials coming soon.

Frequently asked questions

Who is the MI Practice Lab for?
Two audiences. (1) MI trainers and educators who want their workshop cohorts to keep practising — with visibility into how each trainee is progressing. (2) Practitioners who've completed MI training and want to keep their skills sharp. The Lab is purpose-built to complement workshop-based MI learning, not replace it.
How does this complement an MI workshop?
Workshops teach the model — the spirit, the methods, the why. The Practice Lab is what comes after: deliberate, repeated practice with structured feedback. Research consistently shows MI skills decline within months of an initial workshop unless practitioners practise. The Lab fills that gap.
What is the 'share with coach' feature?
After any session, learners can share the full transcript and audio recording with their coach or trainer. The coach gets an interface to leave detailed, inline feedback against specific moments in the conversation — turning the Lab into a between-session check-in tool.
Is the AI's MI scoring accurate?
Early bench testing shows promising alignment between the Lab's MI classification and human MITI-trained coders. A more rigorous evaluation is in progress, and we'll publish results as they're written up. The Lab's scoring is best treated as formative practice feedback, not a credentialing assessment.
Will there be CBT scenarios?
Yes. CBT is on the roadmap and will live inside the same Practice Lab — same interface, same coach-feedback loop, same session structure adapted for CBT. If you train CBT and want early access, get in touch via the For Trainers page.
Can I run this with a cohort?
Yes — this is the most common use case. We work directly with MI trainers running cohort-based programmes. The 'For Trainers' page has more on how cohort visibility, share-with-coach, and bespoke scenarios fit together.