Notes from the front line of AI delivery
What we learn shipping production AI agents inside real enterprises, the numbers, the failure modes, and the delivery model that gets you past the pilot.
- 15AI-Native Enterprise
What Is an AI-Native Enterprise?
An AI-native enterprise is one whose business model breaks if you remove the AI. The removal test, the architectural difference from AI-added, and what it takes.
- 14Forward-Deployed Model
KPMG, EY and PwC all published AI-hallucinated reports in 2026. What it means for who builds your intelligence layer.
Three Big Four firms withdrew or were caught publishing research with fabricated citations in 2026. The failure is structural, and it is a buying signal.
- 13Forward-Deployed Model
Salesforce's own partners are not seeing Agentforce ROI: what that means before you buy a packaged agent platform
Salesforce reports record Agentforce growth while a survey of its own implementation partners finds none crediting it with bookings. What that gap means for a buyer.
- 12Forward-Deployed Model
Accenture, Deloitte and EY All Run Forward Deployed Engineering Now: What Actually Differs
By 2026 Accenture, Deloitte, EY and Salesforce's partner network all brand forward deployed engineering. The question that still separates them: whose vendor funds the pod.
- 11Forward-Deployed Model
Accenture's worst day on the market: what it tells a CIO about the vendor they are about to sign
Accenture had its worst single trading day in June 2026 on soft bookings and AI-driven demand pressure. What that repricing means before you sign a multi-year AI programme.
- 10AI-Native Enterprise
Own Your Intelligence: What It Actually Takes to Build, Not Just Fund
Sequoia, LangChain, Nadella and Karp all say own your intelligence. None of them says who builds it inside a company that is not AI-native. This does.
- 09Forward-Deployed Model
OpenAI's DeployCo, Anthropic's Ode, and the case for an independent forward-deployed partner
The labs now sell forward-deployed delivery themselves. The buying question is no longer who embeds engineers, it is who owns the intelligence when they leave.
- 08Forward-Deployed Model
Build vs buy vs embed: choosing an enterprise AI delivery model
Buy commodity workflows. Build only with a bench you can permanently assign. Embed when the workflow is yours. A decision rule, and where each model breaks.
- 07Pilot to Production
Why 95% of Enterprise AI Pilots Fail (And the 5% That Don't)
MIT found 95% of enterprise AI pilots deliver zero measurable ROI. Here's what separates the 5% that reach production, with 2025-2026 data and a decision-maker's playbook.
- 06ROI & Business Case
The ROI of Enterprise AI Agents: 2026 Benchmarks
2026 ROI benchmarks for enterprise AI agents: self-reported ~171% returns, Klarna's ~$40M, 80%+ resolution rates, and why 95% of pilots still fail. Answer-first, sourced.
- 05Forward-Deployed Model
What Is a Forward-Deployed Engineer (and Why Enterprise AI Needs One)
A forward-deployed engineer embeds in your team, absorbs operational pain, and ships production AI agents on real data under real governance.
- 04Governance & Control
Keeping Humans in the Loop: Human Oversight for Enterprise AI Agents
Human oversight is what separates enterprise AI agents that reach production from the 40% Gartner expects to be canceled. A decision-maker's guide to designing it.
- 03Governance & Control
Audit Trails for AI Agents: What Regulated Enterprises Actually Need
A decision-maker's guide to audit trails for AI agents in regulated enterprises: what to log, why it's a production gate, and how to satisfy the EU AI Act and NIST AI RMF.
- 02Governance & Control
AI-Native Operating Model: Rearchitecting vs. Layering AI on Top
Layering AI onto legacy workflows stalls in pilots. Learn when to rearchitect the operating model for AI-native outcomes, and how enterprise leaders decide.
- 01Governance & Control
Legacy System Modernization With AI Agents: The Mainframe and COBOL Angle
How AI agents change the economics of mainframe and COBOL modernization in 2026, why most pilots still stall, and what enterprise leaders should govern before committing.
The future belongs to those who see it before it is obvious, and build it while everyone else is still out there scouting for the best AI tools.