A Publication on Emerging Software

The Omni Service Application

A new category of software platform is being defined — one that claims to deliver complete outcomes rather than tools to operate. Whether that represents a genuine shift in how software works, or an incremental evolution of existing capabilities, is an open and important question. This publication examines it.


Latest Writing

May 11, 2026 6 min read

Runaway API Loops: The Silent Cost of a Single Bug in an AI Agent

A mismatched variable. A logic error. An unhandled exception. Any one of them can send an AI agent into an infinite loop that burns through your API tokens for hours before anyone notices — if anyone notices at all.

May 6, 2026 7 min read

Vibe Coding and the Illusion of Working Software

AI coding tools can produce applications that look finished and run without errors. That's not the same as software that's secure, maintainable, or correct under real conditions. The difference is what programmers are for.

May 4, 2026 8 min read

Autonomous AI Agents Without Guardrails: The Real Risks to Your Data and Finances

The promise of autonomous AI agents is compelling — software that acts on your behalf across multiple systems simultaneously. The risk, when those agents operate without meaningful constraints, is equally real.

April 29, 2026 7 min read

Legacy Software Modernization: Does OSA Change the Economics?

Hundreds of millions of lines of critical software are running on systems built decades ago. The traditional rebuild options are all bad. OSA platforms claim to change that — but does the claim hold up?

All writing →

About This Publication

The central claim of the OSA category is straightforward: existing software has always required skilled people to operate it. SaaS improved productivity, but humans still did the work. An Omni Service Application, by contrast, is supposed to execute — running across every domain a project requires simultaneously and delivering finished outcomes rather than capabilities to apply.

That's a significant claim, and it deserves honest examination. What does autonomous execution actually mean in practice? Where does genuine delivery end and marketing language begin? What are the real limitations? What do these platforms solve, and what do they not?

This publication doesn't start from the assumption that OSA is the future of software. It starts from the observation that something meaningful is changing in how software is built and delivered — and that the change is worth understanding clearly, without the sales pitch.


Follow the thinking

Occasional writing on the OSA category — what the claims are, what the evidence shows, and what the open questions are. No pitch. No product. Just the ideas.

No spam. Unsubscribe any time.