Do you run Governance Briefings for companies that aren't sure they need AI yet?

Yes. It's a one-day session for senior leadership. You leave with clarity on risks, your local regulatory context, and — above all — what you should NOT automate. We don't sell AI in the briefing — we sell judgment.

Is the Adoption Workshop sold standalone, or bundled with the Sprint?

Both. The Sprint already includes 2 baseline training sessions. The Workshop is for when you already have AI installed (ours or someone else's) and team adoption is low. In-company, 1–2 days, based on teaching patterns instead of tools — methodology with 95% retention.

How long does the Sprint take?

4 to 6 weeks. First working workflow in week 2. Full system by week 6.

Who owns the code and the context?

You do. Everything lives in your repo, your infra, your MCPs. No vendor lock-in.

No. Fixed price per service based on scope. Scope documented before we start.

Why MWP and not LangChain or heavy frameworks?

Because folders are the architecture. It's plain text, versioned in git, interpretable. You understand what the agent does without opening a debugger.

What if we miss the Sprint go-live?

Proportional refund for undelivered scope. Documented in the MSA.

E-Growth Management | The 60/30/10 method for AI in business.

Problem

You bought AI. It never reached operations. 50–75% of enterprise AI adoption never makes it to production.

You tried ChatGPT. Maybe Zapier. Maybe a chatbot. Nothing scaled into daily work. Information across 10 disconnected systems, SOPs trapped in PDFs, tribal knowledge that never makes it into process. You don't know where to start.

You don't need another demo. You need a system that works on Monday.

The method

90% of the work isn't AI.

If you adopt AI without knowing which part of your operation should actually be AI, you fail. 70% of enterprise AI failures come down to people and process — not technology. The 60/30/10 method maps your operation before we touch a model.

60%

Deterministic logic

Traditional code, integrations, straightforward automation. Cheap, fast, doesn't break.

30%

Business rules

Workflows, triggers, rule engines. The known decisions you already make every day.

10%

Real AI

Only where there's judgment, edge cases, or creativity. What no script can solve.

The 60/30/10 Audit tells you exactly how much AI your operation needs. It's almost always less than you think — and cheaper.

Start with the Audit

Services

Five ways to apply the method.

Each service is an entry point and also a step on the path. Start where you are — you don't have to commit to the full sprint on day one.

01
AI Governance Briefing
For leaders unsure whether to jump in at all.
Risk map · local regulatory frame · list of what NOT to automate.
$1.5k–$3k1 day · C-suite session
Book briefing
02Recommended entry
60/30/10 Audit
A map of your operation before we build anything.
What's 60% logic, 30% rules, 10% real AI · full written diagnosis.
$1k–$3k1–2 days · written deliverable
Book audit
03Flagship
MWP Sprint
Full implementation, end-to-end.
System running in production · docs · training · handoff.
$8k–$15k4–6 weeks · fixed price
Book sprint
04
Adoption Workshop
For teams with AI installed but nobody using it.
Team trained on patterns, not tools. 95% adoption methodology.
$1.5k–$4k1–2 days · in-company · up to 30 people
Book workshop
05Recurring
Operations Retainer
Once the system is live and you want it to keep improving.
Ongoing optimization · new skills · new agents · ROI reporting.
$1k–$3k/moMonthly · no lock-in · post-sprint
Let's talk

The sprint

One product. Fixed price. Launch guaranteed.

Context Engineering Sprint. Four to six weeks. $8k–$15k USD based on scope. Documented scope. No hourly surprises.

Price scales with scope: number of MCPs, stack integrations, and departments included. We nail it down together in the discovery call.

WEEK 01

Immersion

Workflow mapping · context audit · architecture.

WEEK 02

Quick win

One workflow live in production. Visible.

WEEK 03–04

Build core

MCPs + skills + RAG + dashboard.

WEEK 05

QA + training

End-to-end testing · session 1 with your team.

WEEK 06

Go-live

Handoff · session 2 · operations runbook.

live

What you take home

agents.md for the business (context, voice, processes)
MWP folder architecture mapped to your workflow
MCPs connected to your stack (Gmail, Calendar, CRM, Notion…)
Skills per department (executable SOPs)
RAG / Knowledge Base with business docs
n8n workflows for core processes
Next.js dashboard with usage metrics
2 training sessions + runbook

How we implement it

Folder structure is the architecture.

No heavy orchestrators. No black boxes. This is a real agents.md, dogfooded in an active client.

A markdown file that humans and LLMs read the same way — that is the MWP method.

— Jake Van Clief · Lyceum

agents.md00_context/

 1# agents.md — Medra
 2
 3> Health tech SaaS, 40+ active doctors, LATAM.
 4> Primary context for every skill in the harness.
 5
 6## Identity
 7- Name: Medra
 8- Category: health tech · medical practice management
 9- Tone: clinical, direct, never marketing-speak
10- Compliance: PHI never in logs. Audit trail on clinical mutations.
11
12## Stakeholders
13- Dr. Julieta · Head of Clinical (clinical validation)
14- Ramiro · Head of Product (skills prioritization)
15- Nico · Ops + harness (stack, automation, this file)
16
17## Core processes
181. Inbound patient triage (5 min SLA)
192. Scheduling + automated reminders
203. PAMI/OSDE billing with audit trail
21
22## Connected tools
23- Supabase + pgvector · Twilio · Resend · HubSpot

context loadedlive

Context file · loaded

agents.md — Medra

Health tech SaaS, 40+ active doctors, LATAM.

Primary context for every skill in the harness.

Identity

Name: Medra
Category: health tech · medical practice management
Tone: clinical, direct, never marketing-speak
Compliance: PHI never in logs. Audit trail on clinical mutations.

Stakeholders

D

Dr. Julieta

Head of Clinical (clinical validation)

R

Ramiro

Head of Product (skills prioritization)

N

Nico

Ops + harness (stack, automation, this file)

Core processes

01Inbound patient triage (5 min SLA)
02Scheduling + automated reminders
03PAMI/OSDE billing with audit trail

Connected tools

Supabase + pgvectorTwilioResendHubSpot

InterpretableDebuggablePortableVersion-controlledPlain text in git. You read the same thing the agent does.

MWP is a formal method. Paper published on arXiv by Jake Van Clief and David McDermott — March 2026.

Direct training: Clief Notes VIP + High Teas · Lyceum Technical Cohort aspirant 2026.

Industries

Same method, applied to your industry.

Metrics, processes, and language specific to each sector. The methodology stays the same — what changes is what we build first and the vocabulary we use.

HealthAI · DentAI Pack

Case: Medra · 40+ doctors in production · LATAM

+25%new appointments
−60%missed calls
5 mintriage SLA

What we automate

24/7 reception and triage
Scheduling with automated reminders
Insurance billing (local standards)
Post-visit follow-up

Talk about your clinic

Cases and benchmarks

Active clients and method benchmarks.

Our own implementations

MWP systems live in production with E-Growth clients.

MedraHealth tech · LATAMin production
Arca YacoWine producer · 40 employees · LATAMin production
InmobiliariaReal estate · LATAMin progress

MWP method benchmarks

Published figures from real MWP implementations — cases from the paper and the wider ecosystem. Not our clients; proof that the framework scales.

451%documented ROIJohns Hopkins revenue cycle.
92%cost reductionPrior auth automation · healthcare · $11.33 → $0.87 per auth.
200+hours savedThompson & Associates · tax season · 3x capacity without hiring.
96%faster responseContract generation · real estate · 14,600 calls with active client.

Read the full paper

Questions

What everyone asks.

Missing yours — ping me on WhatsApp.

The Audit is a 1–2 day diagnosis: it maps which processes in your operation should be traditional code, which should be business rules, and where you actually need AI. It builds nothing — it gives you the map. The Sprint is the implementation of that map. Many clients start with the Audit before deciding whether to commit to a full Sprint.

Before you book, let's see if we're the right fit.

Three questions. If it matches, you book a 30-minute discovery call directly.

How many people are on your team?

Does your company bill in USD or have access to dollars?

What's your main pain point right now?

$8K–$12K · fixed price · no hourly surprises