A read on where AI fits in the 35 Mules playbook. And where a spreadsheet is still the right answer.
Four cohorts in. Twenty plus portfolio companies. A first exit with SustainaBase and ISS-Corporate. A new Gold Coast Tech track running alongside the fifteen-month core program. This page is a short read on what we saw and the first conversation we would like to have.
Most organizations automate the wrong things.
We work on a simple idea. About 60% of what a team does every day is traditional code and database work. About 30% is rule-based logic. Around 10% is a real AI problem. The work is sorting those three into the right layers before anything is built.
Applied to a four-person program running four cohorts, the map resolves quickly. Cohort intake scoring is a rule-based problem. The subject-matter expert calendar is a database and scheduling problem. Portfolio knowledge (who met whom, who needs an intro, which pilot stalled and why) is a structured context problem. The cohort story capture and distillation is where AI actually earns its keep.
Sorting those four into the right layers, before anyone writes a line of code, is the work we would propose to do with the 35 Mules team.
A small team responsible for a fast-moving product. A senior leader who wanted a methodology his team could keep running after we left. The sprint delivered a workshop, advisory calls, an Organizational Context Architecture, and a Strategic Operations Framework.
- Organizational Context Architecture Cohort-ops playbook and portfolio knowledge layer
- Strategic Operations Framework Pilot-to-FPL-business-unit handoff framework
- Advisory calls Through the first full application cycle after close
The methodology is published.
Interpretable Context Methodology: Folder Structure as Agent Architecture, submitted to ACM TiiS. The core idea is that agent context can be organized as a layered filesystem with measurable gains in interpretability and reproducibility. It reads like plumbing because it is plumbing. Open source under MIT license.
Repo: github.com/RinDig/Interpretable-Context-Methodology-ICM-
Companion paper: Ethics Engine (arXiv:2510.11742). A psychometric assessment tool for evaluating ideological and moral patterns in LLMs. Relevant whenever an AI layer has to stand up to outside review.
Who is on the other side of the table.
- 01Jake Van Clief, founder. Marine Corps veteran. MSc Future Governance, University of Edinburgh.
- 02Published in ACM TiiS (ICM) and arXiv (Ethics Engine).
- 031,500 plus enterprise learners trained since May 2025 across Correlation One (Pacific Life, Colgate-Palmolive) and KPMG UK, one of the Big Four.
- 04Jake and Matt also built an online community to 22,000 members in five weeks.
- 05Eduba partners with NLP Logix for work that sits below the orchestration layer. NLP Logix has been in machine learning since 2011 and runs over 150 data scientists.
Thirty minutes with Matt Creamer.
Matt runs enterprise and ecosystem work at Eduba. Bring one workflow from the 35 Mules program you wish you had written down the first time you did it. We will walk through the 60 / 30 / 10 map live and scope the first sprint from there.
If the conversation lands somewhere else, Matt will say so on the call.