EFFEN Applied AI - Executive Summary

EFFEN Applied AI - The Plan (Executive Summary)

An in-house Applied AI function whose single job is lean scaling: grow revenue and output without growing headcount at the same rate. This folder is the shareable plan, in order: theory, roadmap, cost, and the engagement plan.

The thesis, in one line

Net profit impact  =  (labor displaced + revenue lift)  -  (Track A + B + C spend)

As output grows, A+B+C spend grows sub-linearly while the headcount a normal org would need grows linearly. The widening gap is the profit. The cost model proves it: adding a brand costs about 2% of the labor it would otherwise require.

The whole system on one page

flowchart TB
    subgraph SUB["Two substrates (what we build up)"]
      DATA["Data maturity<br/>warehouse, pipelines, MCP"]
      ORG["Organization maturity<br/>tooling, process, AI literacy"]
    end
    K["Shared Context & Knowledge layer<br/>Git + OpenKnowledge vault"]
    subgraph AI["Applied AI (the multiplier, two surfaces)"]
      INT["Internal leverage<br/>reporting, ideation, creative, social"]
      PROD["AI as product<br/>customer DM / CS bot"]
    end
    GOAL["LEAN SCALING<br/>more output, flat headcount, more profit"]
    DATA --> K
    ORG --> K
    K --> INT
    K --> PROD
    INT --> GOAL
    PROD --> GOAL
    classDef g fill:#C6EFCE,stroke:#2E7D32,color:#111;
    classDef k fill:#FFF2CC,stroke:#B7950B,color:#111;
    class GOAL g;
    class K k;

Read bottom-up: the two substrates feed one knowledge layer; both AI surfaces draw from it; both roll up into lean scaling. You cannot out-AI a maturity gap - value is capped by the weaker substrate. That law drives the whole sequence.

The three things we are building

flowchart TB
    ROOT["EFFEN Applied AI - three things to build"]
    ROOT --> P1["1 - Organization maturity"]
    ROOT --> P2["2 - Data maturity"]
    ROOT --> P3["3 - Product / AI"]
    P1 --> P1a["Off WhatsApp onto Notion + Google Workspace"]
    P1 --> P1b["Process captured so agents can run it"]
    P1 --> P1c["Team AI-literate"]
    P2 --> P2a["Warehouse + Airbyte + Dagster on a VPS"]
    P2 --> P2b["Knowledge backbone: Git + OpenKnowledge"]
    P2 --> P2c["Dashboards as conversations via MCP"]
    P3 --> P3a["Reporting, ideation, creative, social"]
    P3 --> P3b["Customer DM / CS bot on our knowledge base"]
    P3 --> P3c["Video, marketplace, CRM later"]
    classDef p fill:#DDEBF7,stroke:#2E75B6,color:#111;
    class P1,P2,P3 p;

The plan, the cost, the ask

• Roadmap: three tracks moving at three tempos - culture (slow, human-paced), infrastructure (phased), product (fast, ship-test-iterate). Reporting Tier 1 ships day one; everything else ships the moment its substrate rung exists. See 02 - The Roadmap.
• Cost: at placeholder numbers, payback in ~5 months, ~RM 3.7M net over 3 years, and each new brand costs a fraction of the labor it replaces. Producing 1,000 ads with AI costs ~RM 1,363/mo vs ~RM 50,000 for the designers it replaces. See 03 - Cost & Unit Economics.
• The ask (today’s onboarding): agree scope + first quick win, align on the role and the part-time -> full-time path, confirm the real numbers, and sign off the (small) budget + access. See 04 - Engagement Plan.

How to read this folder

Doc	What it covers
01 - The Model & Theory	Why it works: two substrates, one multiplier, the gating law, the lean-scaling math
02 - The Roadmap	What ships when: three tracks, phased plan, dependency map
03 - Cost & Unit Economics	The numbers: widening gap, payback, and AI-vs-human per unit
04 - Engagement Plan	Onboarding asks, the 6-month part-time plan, and the full-time scale-up

Full working detail lives in the parent EFFEN folder: Applied AI Context & Operating Model, Applied AI Roadmap, Applied AI Tooling Roadmap, Applied AI Product Build Playbook, and EFFEN_Cost_Model.xlsx.