The concept emerged while balancing consulting, hardware builds, and AI experiments. Every high-performing project shared three traits: obsessive focus, clear metrics, and ruthless automation of repetitive work. The Efficient Company idea captures that operating model so I can refine it with real pilots.
Core Principles
- Automation-first workflows: treat any repeated task as a candidate for scripts, LLM agents, or process redesign.
- Radical transparency: every decision is documented so context survives hand-offs and async collaboration.
- Outcome metrics: define leading KPIs before building; if we cannot measure it, we do not ship it.
- High-leverage people: generalists fluent in engineering, analysis, and storytelling to keep the team tiny.
Operating System Stack
The stack blends custom automations with off-the-shelf services:
- Central knowledge base with living design docs, prompt libraries, and SOPs.
- Model-agnostic prompt framework so we can swap Gemini, Claude, or local LLMs without rework.
- Automation hub (Power Automate, n8n, or custom Python) with observability baked in—logs, retries, alerts.
- Customer-facing dashboards that refresh automatically and surface metrics in plain language for stakeholders.
Experiments to Date
I have applied slices of the framework across consulting sprints and product builds:
- Used modular prompts to power lecture note automation and LeanGPT, proving the process templating approach.
- Ran consulting engagements with templated outreach scripts, reducing prep time while sustaining personalisation.
- Built fabrication projects (Voron 2.4 rebuild, plexiglass sign) with ops checklists that mirror the same structure.
Next Steps
Immediate goals include piloting the model with a small venture studio idea, codifying the SOP library, and stress-testing cost structures against real revenue targets. I am also identifying domains—manufacturing services, AI tooling, specialised consulting—where such a company can deliver outsized leverage.
← Back to article summaries