Search across Doctrine, Practice, Journal, and Community
17 posts
Part three. We did not invent a new SDLC for agents. We put them into the Scrum most teams run, and kept the record: 311 build sessions, $3,268 at list price, 1,470 test runs judged on evidence, and the three things the record caught us getting wrong.
A demo is a foal. It cost nothing and it cannot finish a race. We are a very small shop that breeds the whole horse: data model, application, billing, infrastructure, the console you watch it from, and the support that runs after we leave. Here is the stable, and here is what is still dashed.
A field operator spent years knocking doors, understood the work precisely enough to describe it, and a factory turned that description into a running platform in weeks. Everyone is asking whether AI can build software. The harder question is who knows what to build.
The models were trained on decades of work done the hard way. Nobody has written the rules for working alongside them, because almost nobody has been here long enough to lose anything. We have. Part one of the Stella series.
A client team receives a working enterprise (platform, documentation, support desk, data lake, and a running StellarView each) before any of it is at their risk. The transfer is the last act, not the first.
StellarView draws a codebase as five rings that rotate and pulse. That is for observation, not decoration. Here is the full tour, with the real architecture behind each ring, from GitHub and Confluence.
Mozilla.ai's John Dickerson says control is the new moat, and shipped Otari to prove it. He is right. Here is what that same conviction looks like flying in production: an agent fleet you can see, govern, and cost, frame by frame.
Monaco's John Tracy built a manager agent that spins off intern agents, each with seeded, production-like data. He is right. Here is how we fly it: parallel where parallel makes us accurate, single-agent and sequential where correctness is the whole job.
One build is a proof. A fleet is a practice. StellarView introduced itself with a Gang of Five vertical platforms, governed, documented, and running, with many more waiting in the libraries, demos, and templates.
Two short films StellarView made about itself. Part one follows the method on a real hospital job. Part two walks the running creations.
Everyone talks about AI models. Very few talk about AI systems. Clare Kitching has been saying this for years. StellarView is the proof she is right.
Galaxies, Constellations, Cosmic Emissions, Space Dogs, Wormholes, and the Transporter Room. A funny, honest guide to why a development platform sounds like a NASA mission briefing.
From a 15,000-word specification to 11 executable phases with 78 work units, architecture diagrams, and addressable gaps. The complete breakdown of how Big Bang works.
Every other developer platform assumes your code exists. StellarView creates it, verifies it, deploys it, and manages the client relationship around it. That is the gap.
Skills are firm IP. A CRUD pattern built for one client becomes a launchpad for every future engagement. Inside StellarView's Skill Builder and HAL Template Studio.
Solo Principal managing four concurrent client galaxies on AWS. Ghost instances, Replicant deployments, Construct infrastructure. How Fleet keeps streams from dropping.
Point StellarView at any GitHub repo. In seconds, it maps the architecture into a navigable 3D constellation: five layers, color-coded, clickable. Your codebase has never looked like this.
