EPC Group.net

EPC Group.net Governed Microsoft AI, Data & Cloud — since 1997. Six Microsoft Solutions Partner designations. epcgroup.net

AI governance & vCAIO, Microsoft Copilot, Claude, Power BI, Fabric & Fabric IQ, SharePoint & Teams — AI that holds up in regulated industries. Founded in 1997, EPC Group’s Houston team is a leading Microsoft AI, cloud migration, and Power BI consulting practice. With 29+ years of experience and 11,000+ Microsoft implementations, we help organizations modernize collaboration, analytics, and security using Microsoft 365, Power BI, Microsoft Fabric, SharePoint, Copilot, Azure, and the Power Platform. Our AI consulting and vCAIO services span Copilot, OpenAI, Claude, Gemini, Perplexity, and more. We specialize in tenant‑to‑tenant and M&A consolidations, Exchange and SharePoint emergency support, G‑Suite (Google Workspace) to Microsoft 365 migrations, and legacy file share cleanup, helping Houston‑area enterprises move off outdated systems into a secure, AI‑ready cloud environment.

Microsoft just changed the strategic value of every Power BI semantic model your organization owns.With Microsoft Fabric...
07/28/2026

Microsoft just changed the strategic value of every Power BI semantic model your organization owns.

With Microsoft Fabric IQ, semantic models are no longer only powering dashboards.

They are becoming the trusted grounding layer for answers produced by Microsoft 365 Copilot Chat, Copilot Cowork, Fabric Data Agents, and future AI-powered applications.

That means the measures, KPI definitions, row-level security, sensitivity labels, business terminology, and lineage your teams have built now determine whether enterprise AI answers are trustworthy—or merely convincing.

The central issue with enterprise AI is not whether it can produce an answer.

It is whether the answer is grounded in the right metric, the right definition, the right data, and the right security context.

https://linkedin.com/pulse/microsoft-just-told-you-every-power-bi-semantic-model-have-p3u6e/

In this article, I explain:

• What Microsoft Fabric IQ actually is
• What is changing with Copilot in Power BI
• Why Power BI semantic models are becoming the enterprise AI trust layer
• Why many organizations are not yet ready
• How to certify semantic models for AI grounding
• A practical 90-day Fabric IQ readiness plan

The organizations that spent years governing their semantic layer just discovered that they were not only building business intelligence.

They were building the foundation for trusted enterprise AI.

Is your Power BI semantic layer ready to ground every AI answer your organization produces?

Nearly half of the CFOs we interviewed have never held the one AI governance meeting that matters.Not a hallway conversa...
07/15/2026

Nearly half of the CFOs we interviewed have never held the one AI governance meeting that matters.

Not a hallway conversation. Not a Teams message. Not a slide buried in somebody else’s presentation.

The actual board-visible conversation about AI ambition, autonomy, executive accountability, insurance exposure, governed data, financial controls—and which decisions AI will and will not be permitted to make.

Meanwhile, many of those same organizations are buying Microsoft Copilot licenses, funding agentic AI pilots, expanding Fabric capacity and calling it “AI transformation.”

The CIO can operate the platform. The CDO can govern the data. General Counsel can write the policy.

But only the CFO can convene the meeting that establishes the financial-control boundary.

In this article, I lay out the two-hour executive agenda, EPC Group’s six-pillar AI-readiness framework, the case for multi-model AI on governed data and why the insurance and regulatory pressure is already arriving.

LinkedIn Article: https://www.linkedin.com/pulse/nearly-half-cfos-have-never-held-ai-governance-meeting-errin-o-connor-ynttc/

EPC Group's Full Breakdown:
https://www.epcgroup.net/insights/cfo-ai-governance-conversation-2026

The meeting is two hours. The exposure compounds every quarter it does not happen.

Accenture called it an “isolated matter.”Every CIO should ask a harder question:- Isolated from what?- From Accenture’s ...
07/13/2026

Accenture called it an “isolated matter.”

Every CIO should ask a harder question:

- Isolated from what?
- From Accenture’s operations?
- From client environments?
- From source-code repositories?
- From Azure credentials, storage keys, tokens, or configuration files?

Those are not the same thing.

This article explains the seven questions every board, CIO, CISO, CFO, and General Counsel should ask after a consulting-partner security incident.

An overall framework covering this, read here: https://www.epcgroup.net/insights/delivery-partner-vendor-risk-accenture-2026

We also examine a larger issue:

How much access should any global integrator possess—and who is independently verifying that access?

Read the article below as its a full technical analysis, vendor-risk framework, rate comparison, and 30/60/90-day response plan:

https://www.linkedin.com/pulse/accenture-called-isolated-matter-here-seven-questions-errin-o-connor-hmxic/

Welcome to Subcontractorville... Frontier Company. No, not the airline.On July 2, Microsoft launched the $2.5 billion "F...
07/03/2026

Welcome to Subcontractorville...

Frontier Company. No, not the airline.
On July 2, Microsoft launched the $2.5 billion "Frontier Company" — 6,000 engineers embedded inside Fortune-scale enterprises to deploy AI, with Accenture, Capgemini, EY, KPMG, and PwC pre-named as the delivery tier.
I've seen this movie before. I lived it.

In 2010, a Gartner analyst's phone call put my firm inside Northrop Grumman. That same season, the Federal Reserve Bank of New York hired us to build a Center of Excellence. What happened next — and what happened in the early 2000s, when Microsoft's SharePoint work reached my firm through two layers of subcontracting — is the story behind my take on what Microsoft announced this week.

I call it Subcontractorville. It just got a skyline.

Full story below. 👇
https://www.linkedin.com/pulse/gartner-phone-call-changed-my-firm-microsofts-20th-attempt-o-connor-72lmc/

Six Carriers Told Me the Same Thing in the Same Room. Your Insurance Just (or is just about to) Stop Covering Your AI.I ...
07/01/2026

Six Carriers Told Me the Same Thing in the Same Room. Your Insurance Just (or is just about to) Stop Covering Your AI.

I spent the last 60 days across the table from underwriters at six insurance carriers. Not a panel. The room. This is a new pillar that must be covered.

Every one of them told me the same thing, in their own words: the policy you renewed last January won't cover the AI you've already deployed.

"Silent AI" — the years when your cyber and E&O policies covered AI losses by silence, not by grant — is ending in 2027, and it actually happened this year when the ISO's generative-AI exclusions went live.

Cyber carriers are now capping AI payouts at ~10% of your limit. And the affirmative market that's emerging will only cover you if you can prove you govern your AI.

The receipts are real and costed: Bartz ($1.5B, on data provenance — not training). Moffatt ("no difference whether it comes from a static page or a chatbot"). Mobley (AI hiring tools, certified nationwide). Hasbro ($20M remediation, $40–60M delayed revenue — recovery not automatic).

I wrote the whole thing up — what died, the four-front squeeze, and the six controls the carriers told me they now require before they'll quote.

Your renewal date is your real deadline. Your GC already knows what month that is.



https://linkedin.com/pulse/six-carriers-told-me-same-thing-your-insurance-stop-ai-errin-o-connor-cv6hc

The practice we built to produce the evidence: https://epcgroup.net/services/ai-insurance-readiness | Email: [email protected]

When AI becomes a colleague, the work becomes management.Three full technology cycles. Same pattern every time.Here is t...
06/29/2026

When AI becomes a colleague, the work becomes management.

Three full technology cycles. Same pattern every time.

Here is the full article ↓
https://shorturl.at/LTByi

The marketing promises a finished co-worker. The implementation delivers a half-trained intern. And the enterprise that wins is never the one that bought the loudest tool — it's the one that built the management layer around it.
Right now, that pattern is running again — and almost no one is naming it correctly.

In the last three weeks I watched thirteen separate stories land across enterprise technology that, read side by side, tell exactly one story:

→ AI agent startups are facing a churn cliff because the product was sold as a colleague and shipped as a chatbot.
→ A federal judge just ruled Workday must face a class action over AI hiring bias — "the vendor's model did it" is no longer a defense.
→ Gartner projects that within two years, a single developer's AI token bill will rival that developer's salary. Some users are already burning $20,000 of tokens in a month.
→ Anthropic put a colleague in your Slack channel. NVIDIA just opened the factory to mass-produce more of them — with Cadence, Synopsys, Siemens, Dassault, SAP, ServiceNow, CrowdStrike, and Palantir already on the assembly line.
→ MIT just proved the role boundaries the whole architecture depends on don't exist inside the model. A single technique raises attack success from near-zero to roughly 60% across every frontier model tested.

AI just stopped being a tool you use and became a colleague you hire — and almost no enterprise has built the manager, the payroll discipline, the identity fabric, the performance review, or the termination path required to actually run a workforce that isn't entirely human.

Here is the full article ↓
https://shorturl.at/LTByi

Governance is the floor. Management is the ceiling we forgot to build.

Enterprise AI adoption is accelerating. Enterprise AI governance is not.That gap is becoming one of the largest operatio...
06/24/2026

Enterprise AI adoption is accelerating. Enterprise AI governance is not.

That gap is becoming one of the largest operational, financial, and security risks facing organizations in 2026.

In this new EPC Group executive briefing, Founder and Chief AI Architect Errin O'Connor examines the growing disconnect between AI investment and measurable business outcomes.

Read the full article below.⬇️
• https://shorturl.at/lL4r7

Key findings discussed include:

• 25% of AI initiatives achieving expected ROI.
• Shadow AI usage expanding across enterprise environments.
• Agentic AI systems being deployed without sufficient governance controls.
• Rising concerns around AI operating costs and token consumption.
• The emergence of AI Centers of Excellence as a strategic business requirement.

The organizations that scale AI successfully are not necessarily those with the most advanced models.

They are the organizations with the strongest governance, identity, security, compliance, and operating frameworks.

Read the full article and explore how an AI Center of Excellence can help organizations govern AI at scale.

What AI governance challenges is your organization preparing for in 2026 and beyond?

Read the full article here: https://shorturl.at/lL4r7

🚪 What if the biggest AI risk in your organization isn't the model?🚨 🔗 Read the full article below.👇https://shorturl.at/...
06/22/2026

🚪 What if the biggest AI risk in your organization isn't the model?🚨

🔗 Read the full article below.👇
https://shorturl.at/sXSSE

What if it's the fact that nobody is standing at the front door deciding which models get access to your data?

Over the past few months, I've watched organizations spend millions on AI initiatives while skipping the most important step: governance.
❌ No accountable owner
❌ No model intake process
❌ No AI risk classification
❌ No cross-functional oversight

Meanwhile, AI models are becoming more powerful, more autonomous, and increasingly connected to enterprise data.

The answer isn't another AI tool. It's an AI Center of Excellence.

A properly designed AI COE creates:
🔐 Governance
🎓 Education & Enablement
⚙️ Reusable AI Solutions
🏛️ Executive Steering & Accountability

Before you buy the next AI platform, ask one question:
👉 Who decided that model gets to touch your data?

My latest article explores why every enterprise needs an AI Center of Excellence before it needs another AI tool.

🧠 Everyone's arguing about whether AI is a bubble. They're asking the wrong question.Here's the one that actually matter...
06/19/2026

🧠 Everyone's arguing about whether AI is a bubble. They're asking the wrong question.

Here's the one that actually matters for your enterprise: 👇https://shorturl.at/atbbf

📉 The headlines this month all point one way. A Stanford study found small AI models on a desktop now match or beat the giant data-center models in 80%+ of tasks. Tim Cook says AI's chip appetite makes Apple price hikes "unavoidable." FERC says the grid can't keep up and "standing still is no longer an option."

Translation: raw AI capability is becoming a commodity. Cheap where it's small. Constrained where it's huge. Abundant everywhere.

⚠️ Meanwhile, a quieter failure is unfolding inside enterprises. Gartner: by 2027, 60% of organizations will fail to realize AI value because their data governance and AI governance aren't integrated — calling governance "the single point of failure" for AI ambitions.

💡 Here's what nobody's connecting: these are the SAME story.
When a technology commoditizes, advantage moves UP the stack. When the model gets cheap, owning the best model is worth nothing — everyone has a good-enough one. What becomes scarce, and therefore valuable, is governed data, auditable decisions, regulated-industry trust, and a human accountable for the outcome.

🔑 When AI gets cheap, governance gets expensive. Not wasteful-expensive. Valuable-expensive. Scarce. Worth paying for.

🏛️ That discipline is now the baseline for everyone — healthcare, finance, defense, the mid-market. It's why we built the Governed AI on Microsoft Framework (Purview, Fabric, Power BI, M365, Entra, Copilot, Defender), why we deliver it through the Cloud Orchestrator Practice:

https://www.epcgroup.net/news/microsoft-cloud-orchestrator-practice-launch

🎯Multiple models. One truth.

Full article below and I would love to hear the communities thoughts and comments as it is a really interesting time to be in IT \ AI.
👇https://shorturl.at/atbbf

Address

4900 Woodway Drive Suite 830
San Antonio, TX
77056

Opening Hours

Monday 6am - 6pm
Tuesday 6am - 6pm
Wednesday 6am - 6pm
Thursday 6am - 6pm
Friday 6am - 6pm
Saturday 6am - 6pm

Telephone

+18883819725

Website

, ,

Alerts

Be the first to know and let us send you an email when EPC Group.net posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to EPC Group.net:

Shortcuts

Share

Category