Insights

Original writing on governed AI, enterprise knowledge, expertise, leadership, and invention.

By Christopher Jackson, founder of Corvair.ai. Official blog on Medium · Published on e27

Official blog on Medium

Why Your AI Workflow Works in the Demo and Falls Apart in Production

July 23, 2026

Why Your AI Workflow Works in the Demo and Falls Apart in Production

software-engineering multi-agent-systems ai-reliability

Why does a multi-step AI workflow that looked flawless in testing start failing constantly once it’s handling real volume? The answer usually isn’t that the agent got worse. It’s math. Specifically, it’s what happens when you chain steps together....

Six Sigma Came for Manufacturing. Now It’s Coming for AI

July 21, 2026

Six Sigma Came for Manufacturing. Now It’s Coming for AI

ai-governance machine-learning agentic-ai

Every regulated industry has a way to measure quality. Banking has capital ratios. Manufacturing counts defects per million parts. Aviation tracks incidents per flight hour. Ask any of them how they’re doing, and they can give you a number. Ask an...

Why Agentic AI Needs More Than Standard Model Risk Management

July 19, 2026

Why Agentic AI Needs More Than Standard Model Risk Management

agentic-ai artificial-intelligence banking-regulation

Here’s a question worth sitting with: what happens when the thing you’re supposed to be governing won’t hold still long enough for you to govern it? That’s not a rhetorical trick. It’s the actual problem underneath agentic AI risk. Model risk...

Inside the Agentic AI Regulatory Matrix

July 17, 2026

Inside the Agentic AI Regulatory Matrix

regtech agentic-ai risk-management

If your agent broke a rule tomorrow, which Regulator would object? Sit with that for a second. Most people assume the answer is “plenty.” Surely between the EU AI Act, GDPR, SR 11–7, and a dozen other frameworks, agentic AI is covered from every...

Budgeting for AI Agents

July 12, 2026

Budgeting for AI Agents

ai-governance ai-transformation ai-economics

Traditional IT budgeting assumes cost predictability. You estimate next year’s infrastructure based on historical growth, negotiate multi-year contracts, and build in incremental adjustments. This works when costs scale linearly with user headcount...

Total Cost of Ownership for AI Agents

July 10, 2026

Total Cost of Ownership for AI Agents

ai-economics ai-transformation ai-governance

Most TCO models for AI agents cover three cost layers: build, infrastructure, and model consumption. These are the obvious ones, and they’re straightforward to estimate. The challenge is that they represent roughly 40 to 60 percent of the actual...

Token Economics: The Operational Foundation for Agentic AI

July 08, 2026

Token Economics: The Operational Foundation for Agentic AI

ai-governance agentic-ai ai-economics

Tokens are the unit of AI economics. Every decision an agent makes, every reasoning step it takes, every mistake it retries translates directly into cost. Understanding token consumption patterns is what separates sustainable agent deployments from...

The Risk Economics of Agentic AI

July 06, 2026

The Risk Economics of Agentic AI

ai-transformation agentic-ai ai-economics

One number to start with: the most comprehensive regulatory framework available today covers roughly 14 percent of the identified risks in agentic AI deployments. Corvair’s Agentic AI Risk Catalog identifies 133 distinct risks across 15 categories....

Building the Business Case for an AI Agent

July 03, 2026

Building the Business Case for an AI Agent

ai-strategy ai-economics ai-governance

At some point, every agentic AI initiative needs to answer the same question: should we invest in this, and if so, how much? Traditional IT business cases aren’t quite the right vehicle. They assume static costs, linear scaling, and risk profiles...

The Economic Case for Agentic AI

July 01, 2026

The Economic Case for Agentic AI

ai-governance ai-transformation agentic-ai

Most organizations building business cases for AI agents are using a software economics model. Estimate the compute, divide by headcount, and compare to hiring costs. It’s a reasonable starting point, but it misses something important about how...

The Great Agentic Capture

December 16, 2025

The Great Agentic Capture

Over the course of this series, we dissected the “definitive guides” from ten of the most powerful technology firms on earth. We waded through white papers, technical architectures, and executive playbooks released over the last six months. Each...

The Pipeline, the Paycheck, and the Petri Dish

December 11, 2025

The Pipeline, the Paycheck, and the Petri Dish

There is a pattern hiding in plain sight in the AI industry, and once you see it, you cannot unsee it. Every company talking about “AI Agents” is not actually talking about agents at all. They are describing what they already sell while wearing a...

The “Cog” and the “Code”

December 11, 2025

The “Cog” and the “Code”

The deeper we dig into the “Agentic AI” narrative, the clearer the pattern becomes. Every vendor is rushing to define “Agents” in a way that conveniently validates their existing product roadmap. We’ve seen Agents defined as queries, microservices,...

The “HR” Fallacy and the “Headcount” Trap

December 09, 2025

The “HR” Fallacy and the “Headcount” Trap

In my previous posts, we debunked the notion that “Agentic AI” is just a new term for old workflows or database queries. Now, we turn our attention to the Administrators. I just reviewed two new reports: “The Agentic AI Advantage” (Dataiku) and “AI...

The “Agentic” Shell Game

December 07, 2025

The “Agentic” Shell Game

We need to talk about “Agentic Orchestration.” The narrative is consistent: AI Agents are the future, but they can be unpredictable and high-risk. The proposed solution? Orchestration. It sounds logical. We need a conductor for the orchestra. We...

The “Agentic” Masquerade

December 07, 2025

The “Agentic” Masquerade

The automation industry is currently engaged in a frantic game of dress-up. In my last post, we examined how legacy workflow platforms are scrambling to rebrand their rigid, deterministic flowcharts as Agentic Orchestration. They are taking old...

Published on e27

Christopher Jackson’s articles on AI, judgment, leadership, and invention. Full articles are published by e27. Explore IPGuru.ai for idea refinement and prior-art research.

September 07, 2026

AI has answers, experience has judgment

AI makes information and creation cheaper, increasing the value of domain expertise, practical experience, and judgment in applying ideas.

August 31, 2026

SWOT is not boring; you are just using it too late

5W1H, SWOT and PESTLE help innovators test problems, assess risks and improve ideas before investing significant time or money.

August 24, 2026

SEA has the ideas; it needs the follow-through

Southeast Asia’s innovation challenge is turning practical ideas into assets through stronger pathways for testing, protection and follow-through.

August 17, 2026

AI is not the advantage, build what competitors cannot copy

AI can help Southeast Asian SMEs turn practical know-how into defensible products, processes and intellectual property that rivals cannot easily copy.

August 10, 2026

The most dangerous place for a good idea is your head

AI can help Southeast Asians turn practical ideas into protected inventions that create lasting business value.

June 09, 2026

The job description is lying to you, and it’s costing you your best hires

AI exposes the limits of static job descriptions, shifting hiring toward judgment, context, and real competencies.

June 03, 2026

Great talent used to mean getting things done, now it means knowing what to get done

AI makes execution abundant, increasing the value of judgment, clarity, and the ability to shape systems.

May 29, 2026

If you’ve become irreplaceable, you’re the problem

AI is turning leadership taste into an organisational bottleneck unless judgment becomes visible and teachable.

Follow Us on Medium

Follow the official Corvair blog for new writing on governed AI and enterprise knowledge.

Follow on Medium

Want to Discuss What You've Read?

Schedule a complimentary briefing to explore how these ideas apply to your institution.

Schedule a Briefing