Memos — Ampron

№14 13 July 2026 Published

AmpronLED API Quick Start Guide

From unboxing to API-controlled — network setup, SSH access, and the essential HTTP endpoints

Every Ampron LED and LCD display ships with a plain HTTP API on board — no proprietary controller, no vendor software, no licence server. This guide covers the factory-default network settings, how to move a display onto your own network safely, SSH file access, config.json, and the handful of endpoints that handle control, configuration, and debugging. Written for integrators and technical teams commissioning their first Ampron display.

№12 14 June 2026 Published

Display as an Integration Layer

How to add reliable display output to a system you've already built — without taking on a display platform

Every display vendor wants to become a platform. What most integration projects need is simpler: HTTP in, content on screen. This memo is written for software integrators and project managers who have a working system and need a display layer — not a platform dependency. Covers the API pattern, push vs pull, offline behaviour, hardware specs, and three concrete use cases.

№11 9 June 2026 Published

Why Multiple Specialised Agents

On why specialisation matters when AI does real work, even though the underlying model is the same

If a frontier AI model already knows everything, why build a team of specialised agents instead of one generalist? The answer is not about knowledge — it is about what surrounds it. Seven reasons: focus, scope and tools, how role shapes reasoning, continuity per domain, parallelism, failure isolation, and clean handoffs to humans.

№09 22 May 2026 Published

Agents Under the Executive

How Bezos's three-decisions-a-day principle should shape the design of AI agents in leadership functions

Bezos's 'three decisions a day' claim is not a productivity tip — it is a statement about what executive labour is for. This memo translates that principle into a design framework for AI agents in leadership functions: accountability as the ceiling, reversibility as a system property, compounding error as the underappreciated failure mode, and four lean documents as the governance standard. Eight sections, one clear argument.

№06 17 May 2026 Published

Selling Ampron — A Starter Guide

A product, API and lead-finding guide for new sales associates

What Ampron makes, where to send a customer for detail, how the REST API works in plain language, who we sell to, and how to find leads — with AI prompts to speed the search.

№03 12 May 2026 Published

Ampron Memos

An introduction to the page you're reading

What Ampron Memos is, what it will hold, and why a bookmark here will keep paying off — for integrators, partners, customers' engineers, and anyone sizing up how Ampron works.

№01 1 May 2026 Published

The Agent Executive Team

A Strategic Vision for the First Generation of AI-Native Executives

The highest-leverage use of frontier AI is not at the task layer — it is at the director tier. This book makes the case for constituting a structured team of AI agents, each holding a real executive domain, and running them with the discipline you would apply to a human C-suite. Twenty chapters on the architecture, the charters, the failure modes, and the strategic logic for doing this now.