Tag

Agentic AI

AI Automation

Nine Production Automations Have 31 Endings. Fewer Than Half Are the Happy Path

Almost nobody publishes the architecture of the automations they deliver, so almost nobody can say what one actually looks like. We publish nine, which means we can count them. Across 139 nodes and 148 edges in nine sectors, the systems fan out rather than funnel: 15 ways in against 31 ways out. And of those 31 endings, only 15 are the thing the system was built to do. The rest are refusals, holds, escalations and self-maintenance.

IrenicTech · Sep 23, 2026 · 7 min read

SaaS Architecture

Agent-Ready SaaS: Building Software Other People's Agents Can Use

For twenty years the question was whether a person could use your product. The new question is whether an agent can, because agents now shortlist, evaluate, operate and increasingly buy software on someone's behalf. Agent-readiness is not a feature you add later; it is a property of your data model, your API and your permission system. This is what it actually requires.

IrenicTech · Sep 21, 2026 · 8 min read

Agentic AI

AI Agent Development in 2026: From First Prototype to Production

Most writing about AI agents is either a demo or a warning. This is the working guide: what an agent is made of, how it reaches your systems through MCP, when one agent should become several, how much the model choice actually matters, why the majority never reach production, and what governance has to be in place before any of it is load-bearing. Every section links to the deeper piece on that question.

IrenicTech · Sep 21, 2026 · 16 min read

AI Automation

We Have 21 Decision Points in Production. Jev Would Change Six of Them

TypeSafe AI released a model that cannot hold a conversation, which is the point of it. Jev takes unstructured state in and returns a typed decision with a calibrated probability, in 70ms to 500ms. We went through the 21 decision nodes in our published automation teardowns to see where a model like this actually belongs. The answer is six of them, and the reasoning matters more than the number.

IrenicTech · Sep 21, 2026 · 7 min read

Agentic AI

Developers Use AI for 60% of Their Work and Can Delegate Almost None of It

Anthropic's own research says developers use AI across roughly 60% of their work but can fully delegate only 0% to 20% of tasks. That gap is not a failure of the models. It is the actual shape of agentic development, and it changes what you should expect from an AI-native team.

IrenicTech · Jul 12, 2026 · 5 min read

E-commerce

AI Agents Are Buying Things Now. Here Is What Your Store Has to Implement

Agentic commerce stopped being a demo. Three protocols now cover discovery, payment, and authorization, and as of June 2026 Shopify lets any developer register an agent without applying. The merchant-side work is concrete, and most stores have not started.

IrenicTech · Jul 12, 2026 · 6 min read

DevOps

Agents Don't Fix Broken Engineering Practices. They Scale Them

The best line written about agentic development this year is that agents do not magically fix broken engineering practices, they scale them. If your tests are thin and your review is a rubber stamp, an agent makes that worse at speed. Here's the readiness check.

IrenicTech · Jul 12, 2026 · 6 min read

Agentic AI

What Is Agentic AI Development in 2026? Orchestration, MCP, and Multi-Agent Teams Explained

Agentic AI development moved from autocomplete to orchestration in 2026. Here's what an AI agent actually is, where MCP and multi-agent teams fit, and what it means for building real software, without the hype.

IrenicTech · Jun 28, 2026 · 6 min read

Agentic AI

Agentic AI Security: Governing AI Agents Before They Become Your Next Breach

An AI agent is an autonomous user with access to your systems, and that's exactly what makes it a security problem. Here's where agentic AI gets breached, why analysts expect it to drive real damage by 2028, and how to govern agents before it's your incident.

IrenicTech · Jun 28, 2026 · 5 min read

Build vs Buy

Build vs Buy AI in 2026: When to Use an Off-the-Shelf Agent and When to Build Your Own

Most companies don't need to build AI from scratch, and some can't afford not to. Here's a clear-eyed framework for deciding when an off-the-shelf agent is enough and when a custom build is the only thing that actually fits.

IrenicTech · Jun 28, 2026 · 4 min read

Agentic AI

Dubai's Agentic AI Plan: What It Means for Businesses in the UAE

Dubai is pushing agentic AI into its private sector with a two-year plan to become the world's leading city for commercial AI adoption. Here's what that means for UAE businesses and how to prepare without the hype.

IrenicTech · Jun 28, 2026 · 4 min read

Multi-Agent Systems

Multi-Agent Systems: How Teams of AI Agents Are Starting to Run Software Work in 2026

One AI agent is useful. A coordinated team of them is a different kind of power, and a different kind of risk. Here's what multi-agent systems actually are, where they earn their keep, and what it takes to run them without creating chaos.

IrenicTech · Jun 28, 2026 · 4 min read

MCP

MCP Explained: The Protocol That Makes AI Agents Useful, and How to Govern It

MCP is the standard that lets AI agents actually use your tools and data instead of just talking about them. Here's what it is, why it took over in 2026, and why connecting agents to your systems is now a governance problem too.

IrenicTech · Jun 28, 2026 · 4 min read

MCP

Is Your SaaS Agent-Ready? Why MCP Is the Integration You Can't Skip in 2026

AI agents are starting to choose and use software on their own. The SaaS that exposes an MCP server becomes the one they can actually use, and that's the integration winning enterprise deals in 2026.

IrenicTech · Jun 20, 2026 · 4 min read

Agentic AI

Most AI Agents Never Reach Production. Here's What Separates the Ones That Do

Pilots are easy; production is where AI agents go to die. Here's why most never make it, and the unglamorous engineering, evals, guardrails, feedback loops, that separates the agents that ship from the ones that stall.

IrenicTech · Jun 13, 2026 · 3 min read

Claude Fable 5

Claude Fable 5: Anthropic's Most Capable Public Model Is Here

Anthropic's Claude Fable 5 is its most capable public model yet. Here's what shipped, the new fallback safeguards, pricing, and what it means if you're building AI products.

IrenicTech · Jun 11, 2026 · 4 min read