Skip to content

AI PROJECT · COMMERCIAL INTELLIGENCE FRAMEWORK

The Project Behind
MAYA OS

An overview of the framework and steps MAYA OS uses to diagnose business problems before prescribing a fix. For the full experience and how to get started, visit the MAYA OS page directly.

Project status

Evolving Prototype

Business problem

Organisations already hold marketing, sales, and CRM data, but it sits in separate systems measured on metrics that do not agree. Meetings end in opinion rather than an auditable conclusion.

Why the problem matters

When nobody can see which stage is losing revenue, more advertising spend becomes the default answer — and the original problem is scaled up rather than fixed.

Who this is for

Designed for executives, business owners and teams across Marketing, Sales, CRM and Data who need a structured view of business problems before deciding what to address first.

Project purpose

Build a framework that forces diagnosis before prescription, with AI doing analysis and hypothesis generation while a human makes the decision.

Chanikan's role

Designed all five stages, the diagnostic question structure, and the human review gates; tested them against a self-created fictional dataset.

Tools and technology

  • A large language model for analysis and drafting
  • Google Sheets for the fictional dataset
  • Self-designed prompt architecture templates
  • A written review log

Workflow overview

Work Requirement
Input
AI-Assisted Processing
Accountable Review
Practical Output

Workflow

  1. Collect & Connect Data — gather what exists and normalise it into one measurement language
  2. AI Analysis — surface relationships, anomalies, and hypotheses worth testing
  3. Human Review — test hypotheses against real context and discard the irrelevant ones
  4. Action Plan — convert conclusions into sequenced priorities, owners, and measures
  5. Review Rhythm — set the cadence that keeps decisions continuous

Expected output

  • A diagnosis report naming the stage where revenue leaks
  • A hypothesis set ranked by likely impact
  • A sequenced priority plan with named owners

Example application

Illustrative workflow example

A fictional revenue-leak diagnosis example

Input

  • Fictional marketing and sales data
  • Metrics from multiple systems
  • Team meeting notes

AI-assisted process

  • Normalise data into one measurement language
  • Surface relationships and anomalies
  • Hypothesise which stage is leaking revenue

Human Review

  • Test hypotheses against real context
  • Discard the irrelevant ones
  • Record the reasoning for later audit

Output

  • A sequenced priority plan with named owners and measures

Review by the accountable owner

AI never closes the loop on its own. Every hypothesis passes through the stage-three human review, where the reviewer must record why each one was accepted or rejected so the reasoning can be audited later.

Responsible AI and data boundaries

  • Examples use public, sample, or synthetic data
  • Customer data, personal data, or confidential data should never be entered into a public demo
  • AI output must be reviewed before it is used
  • This project demonstrates an approach — it is not a fully autonomous enterprise system

Current limitations

  • No paying client deployments
  • Tested only against fictional datasets; not validated at scale
  • No automated integration with CRM or advertising platforms
  • Output quality depends entirely on input data quality

Next development stage

Turn the diagnosis steps into reusable templates and test the framework in a small-group trial round using fictional data.

Chanikan's methodology

Chanikan Kanchanasalee

AI-Native Business Strategist specialising in Commercial Growth

Experienced in leading cross-functional work across Commercial Growth, Marketing, Sales, CRM, Data and Operations.

Related project

Explore further