RMIT School of Psychology · AI Training Case Study · March 2026
When AI stops being a thing you ask for answers and starts being a colleague you think alongside, the work changes. Over eight weeks, the RMIT School of Psychology made exactly that shift — and recovered the cost of the program almost twice over in the first year.
Snapshot
- Client: RMIT University, School of Psychology
- Duration: 8 weeks (5 February – 26 March 2026)
- Format: 8 × 4-hour hands-on workshops
- Investment: $12,800
- Core adopters: 3–5 researchers attending 5+ sessions each, across 31 academics with access
- Legacy materials: 200+ pages of reusable research briefs, handouts, decks and transcripts
I used to use AI just to generate stuff, search, or do monotonous things. Now I see it almost like another colleague — someone you can bounce ideas off, evaluate with, and error-check against. That’s been the biggest enlightenment for me.
— Russell Conduit, Associate Dean (Psychology), RMIT University (attended 7 of 8 workshops)
The short version
RMIT’s psychologists weren’t short of AI exposure — they were short of AI that fit their actual work. University-wide sessions covered prompting basics, but nothing connected AI to psychology research methods, and the school’s 31 academics had scattered, inconsistent relationships with the tools. Some trusted AI blindly; others avoided it entirely. Almost no one was using it inside a real research workflow.
We ran eight 4-hour, hands-on workshops built entirely around real psychology tasks — real transcripts, real ARC criteria, real PICO frameworks — and led with method before tool. We showed the failure modes on purpose: live citation errors, sanitised emotion in qualitative data, automation bias. Trust was built through transparency, not hype.
The result was a measurable shift from “generation tool” to “colleague.” Data-extraction work that had taken eight months was reproduced in twenty minutes. A 500-paper screening ran with zero errors across 200 spot-checked papers. Core adopters began training their own students. On conservative numbers, the engagement returned 195%+ of its cost in year one — before counting the long tail of 200+ pages of materials available to academics who never set foot in a session.
The challenge
- No domain-specific training — university AI sessions covered prompting basics, not psychology research methods.
- No verification culture — researchers either trusted AI blindly or avoided it entirely.
- No methodology integration — AI was used as a search engine, not embedded in research workflows.
- 31 academics with scattered, inconsistent AI relationships across the school.
What we built
| Deliverable | Outcome |
|---|---|
| 8 hands-on workshops | The full arc from mindset shift to reproducible research pipelines |
| 11 research briefs | 50,000+ words of peer-reviewed evidence synthesised for the school |
| 8 slide decks & 8 exercise handouts | Plus 7 clean transcripts and prompt-template packs — 200+ reusable pages |
| The ResearchOps Loop | A unified methodology: Ingest → Structure → Transform → Verify → Ship |
The turnkey difference
- Domain-specific, not generic. Every exercise used real psychology research tasks, so researchers integrated AI into methodologies they’d already mastered — not toy examples.
- Methodology-first, not tool-first. We started with “here’s what your research needs,” not “here’s what the AI can do.” Epistemological limits came first; practical application followed.
- Honest about failures. We demonstrated 20% citation-error rates live, showed emotion sanitisation in qualitative data, and triggered automation bias on purpose — building trust through transparency.
- Materials as institutional legacy. 200+ pages of briefs, handouts, decks and transcripts are available to all 31 academics, including those who didn’t attend. The value extends years beyond the program.
Investment & return
| Category | Item | Value |
|---|---|---|
| Investment | RMAI engagement | $12,800 |
| Year 1 value | Recovered researcher time (year-to-minutes extraction, zero-error screening, reusable materials) | 195%+ of fee |
| Year 1 recovery | Conservative | 195%+ |
| Ongoing | Teaching multiplier (core adopters now training their own students) | Compounding |
| Shift from tool to colleague | Priceless |
In their words
What’s blowing my mind is how rapidly this is evolving. I had a look four or five years ago and it’s so different now. Back then it would just hallucinate gibberish; now it’s a trusted partner.
— Russell Conduit, Associate Dean (Psychology), RMIT University
I uploaded all 500 PDFs into Claude and got it to look for screening for sleep disorders and excessive daytime sleepiness. I checked about 200 of the 500 papers and hadn’t found a single mistake.
— Russell Conduit, Associate Dean (Psychology), RMIT University
Talk about looking a gift horse in the mouth. Staff should be taking advantage of this — I don’t know where else you’d get this level of introduction to AI.
— Russell Conduit, Associate Dean (Psychology), RMIT University
I started teaching AI to my students. I did it this morning. They were so amazed. I taught them where to use it and how to use it — because if you don’t do that, they learn it the wrong way and get AI to write for them without criticising.
— Leila Karimi, Professor, RMIT University
One key reason I wanted to come here is so that we as a department are on the same page about AI. Discussing it together breaks down barriers and increases our understanding.
— Marcel Takac, Lecturer, RMIT University
What made it work
The pattern is repeatable: meet experts inside their own domain, teach the method before the tool, and be honest about where AI fails. People adopt AI as a colleague when they’ve seen exactly where it can — and can’t — be trusted. That’s not a lecture about AI. It’s practice with their own work, alongside someone who has done it.
Want this for your team? Talk to Real Minds AI.