— Methodology
Intuition asks.
AI answers.
Most AI projects fail because they automate the wrong question. We work the other way around — start with the operator's intuition, ground it in domain expertise, and let AI run it at scale.
— The framework
Three legs of a stool.
Remove one and it falls.
We don't believe in AI alone. We don't believe in human-only either. The work that survives reality combines three forces, deliberately.
The operator's gut
Pattern-matching on lived experience.
Every business owner already knows things their software doesn't — what's actually broken, who's actually doing the work, where the real bottleneck sits. Intuition is the right question, before you even know how to phrase it.
Machine memory
Prediction, recall, scale.
AI is what listens, remembers, and runs the answer at scale. Not the magic. Not the goal. The multiplier on judgment that's already there. Used well, it removes friction; used badly, it produces the wrong answer faster than anyone can catch it.
Local depth
Knowing how the work survives reality.
Cambodia tax compliance. Multi-country budget reconciliation. KHQR payment plumbing. Khmer-language UI. The thousand details that decide whether the system works on day 91. AI without this gets it 90% right and breaks on the edge cases that matter.
— What goes wrong
Two legs
isn't enough.
The teams we see fail are the ones running on two of the three. Each combination produces its own predictable failure mode.
Intuition + AI
(no domain depth)
Clever output that breaks on edge cases
The pilot demos beautifully. Then it hits a Cambodia tax bracket, a multi-currency consolidation, or a language quirk — and the AI confidently produces nonsense.
Intuition + Expertise
(no AI)
Humans alone, no scale
The right people doing the right things, slowly. Decisions take days. Knowledge stays in heads. Growth caps at the founder's calendar.
AI + Expertise
(no intuition)
The wrong question, answered fast
Technically precise. Operationally useless. The team automates what's easy to measure, not what actually matters.
— What it sounds like
"Anyone can deploy AI. Few can ask the right question. Fewer still can make the answer survive contact with reality."
— Service map
How the framework
shows up in practice.
Every service we offer leans on the three legs differently. Here's how to read it.
Odoo ERP Implementation
Intuition · Domain · AI (where useful)
Your operators tell us how the business actually runs. We build the ERP around that, leaning on Cambodia and multi-country domain depth. AI shows up where it removes friction — invoice extraction, anomaly detection, customer service.
Learn more →Fractional Leadership
Intuition · Domain · AI
Senior judgment from operators who've run companies, paired with domain expertise across tech, finance, and operations, paired with AI tooling that gives the executive an honest dashboard instead of a curated one.
Learn more →Cambodia Accounting & Tax
Domain (heavy) · Intuition · AI (selective)
The most domain-heavy service we run. NSSF, GDT e-invoicing, dual currency, withholding tax brackets — the rules are complex and change. AI helps with classification and reconciliation; intuition catches what doesn't smell right.
Learn more →Rapid Prototype Studio
Intuition (heavy) · AI · Domain
Where we ship our own ideas in weeks. Intuition leads — what should exist? Domain expertise picks the right tools. AI accelerates the build. Better Meetings, KHQR Bot, and our community tools all came from this loop.
Learn more →— Working together
Building something that needs all three?
A 30-minute call to understand your business, your current systems, and where the framework would land for you. No commitment.