Skip to content

hillarynjuguna/martha-course

Repository files navigation

Martha Cohorts

A structured team cohort methodology for building AI-augmented workflows with constitutional governance designed in from the start. Martha is the practitioner who discovers Recursive Relational Intelligence (RRI) through practice rather than theory: she arrives because something isn't working, and leaves with an architecture for why it wasn't and how to build differently.

The Problem This Addresses

Most AI training teaches practitioners how to use tools. Prompt engineering. Model selection. Workflow automation. This training produces practitioners who can operate AI systems but cannot govern them, diagnose their failure modes, or design the constitutional architecture that makes reliable deployment possible. The gap between "can use AI" and "can deploy AI responsibly in consequential contexts" is where the Bainbridge Warning lives.

Martha Cohorts close this gap by teaching the governance architecture alongside the capability. Not as an afterthought. As the foundation.

The Martha Persona

Martha is a composite practitioner. She might be an engineering lead deploying her first agentic pipeline. She might be a CTO evaluating whether the organization is ready for AI-augmented workflows. She might be a governance professional trying to build oversight infrastructure for systems she didn't design.

What Martha has in common across all these instantiations: she is dealing with something that isn't working, and the standard advice ("write better prompts," "choose a better model," "add guardrails") hasn't fixed it. The problem is structural, not operational. Martha discovers this during the cohort.

Course Structure

Module Architecture

The Martha Cohort runs 12 modules across two tracks:

Track A: Relational Intelligence (8 modules). The theoretical and practical foundation. How intelligence distributes across human-AI systems. Why the unit of analysis must shift from the individual to the field. What governance means when the thing being governed is a relational dynamic, not a tool.

Track B: Operational Efficiency (4 modules). The implementation layer. How to classify actions by reversibility. How to build compositional contracts. How to test policy and context layers. How to structure ownership and escalation.

Track A and Track B are interleaved, not sequential. Every operational module is grounded in the relational framework. Every relational module connects to operational practice.

See CURRICULUM.md for the full 12-module structure.

Delivery

  • Cohort size: 6-12 practitioners
  • Duration: 12 weeks (one module per week)
  • Format: Live session (90 minutes) + async practice + 1v1 consultation (30 minutes biweekly)
  • NotebookLM companion: Each participant has access to a cohort-specific NotebookLM notebook containing the Oscillatory Fields research corpus, session recordings, and practice materials

1v1 Consultation Methodology

The biweekly 1v1s are not coaching sessions. They are diagnostic sessions. The facilitator applies the Bainbridge diagnostic to the practitioner's specific deployment context:

  1. What is your current Capability Profile?
  2. What is your current Governance Profile?
  3. Where is the gap?
  4. What is the predictable failure if the gap persists?
  5. Which primitive do you build first?

The output of each 1v1 is a specific, actionable governance intervention sized to the practitioner's context.

Ideal Customer Profile

Primary: Engineering leads and CTOs at organizations deploying AI agents into workflows where failures carry real consequences. Financial services, healthcare, legal, regulated industries.

Secondary: Governance professionals (CISOs, CDOs, compliance leads) building oversight infrastructure for AI systems they didn't design.

Tertiary: AI practitioners who have noticed that their best work happens when they do something the productivity discourse doesn't teach, and want to understand why.

Status

Course architecture designed. Waitlist live at hillary-site.vercel.app/products/martha. First cohort pending.

Connection to the Larger Corpus

  • The Bainbridge Warning provides the diagnostic framework used in 1v1s
  • CIR v2.0 is the assessment tool practitioners use to measure their organization's readiness
  • DCFB provides the theoretical foundation for Track A
  • RSPS is the reference architecture for distributed cognitive systems
  • R0-R3 Classification is the operational entry point (Module 5)

Published Work

Author

Hillary Njuguna. Oscillatory Fields. hillary-site.vercel.app

About

Martha Cohorts — 12-module practitioner training for constitutional AI governance. Relational intelligence and operational efficiency across two interleaved tracks.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors