Emergent Software

05/28/2026

Enterprise AI problems are often data access problems in disguise.

When different systems access data in different ways, trust drops, governance becomes harder, and scaling becomes far more complicated.

This carousel highlights a few common questions around unified data access layers and AI architecture.

You can read the full blog by Principal Software Architect Aaron Varga here: https://vist.ly/55sh6.

05/27/2026

AI shows up in our development work in two ways. How we write code, and how we build solutions for clients.

Mark Bajema and Mike Allen talk through what that actually looks like day to day at Emergent Software, where AI speeds things up, where it falls short, and why developers are still very much in the driver's seat.

05/26/2026

Is your organization ready for AI agents at scale?

Korey Stillman walks through the numbers in this clip. Microsoft projects 1.3 billion agents in use by 2028. 82% of organizations plan to integrate agents within one to three years. 40% plan to start this year.

The harder question is what happens next. Visibility into which agents exist, what they're accessing, and how they're governed becomes a foundational IT concern, not a future one.

His full blog covers Microsoft 365 E7, frontier workers, and why identity is the control plane for AI-enabled environments: https://vist.ly/55hsw.

05/25/2026

Today, we remember the men and women who gave their lives in service to our country.

Thank you.

05/22/2026

3,000 patients. 1,500+ automated orders every month.

That's the usage on the patient portal we built for Pediatric Home Service, a Minnesota provider of medical equipment and home care for children with complex medical needs.

We rebuilt their portal and integrated it with BrightTree so families can log in, browse supplies, and submit orders that flow directly into PHS's system. Our ongoing work also covers Azure, SQL, Microsoft licensing, and reporting.

Read the full review on our Clutch page: https://vist.ly/556v9.

05/21/2026

In his latest blog, Aaron Varga points out that roughly 80% of the work required to move AI into production is data and access work rather than model work.

That's why switching AI vendors rarely fixes underperforming initiatives. Every model eventually runs into the same enterprise environment, and that environment is usually where the real complexity lives.

The post covers why organizations underestimate the operational side of AI, what fragmented data access looks like in practice, and how things change when teams build for reusability from the start.

Check out his blog here: https://vist.ly/54zwy

05/20/2026

There's one question Aaron Varga says tends to show up when AI architecture starts breaking down: "Which answer is the right one?"

When two AI systems return different answers to the same question because each is interacting with a slightly different version of the underlying data, trust erodes quickly. And once trust erodes, adoption usually follows.

His latest blog covers why inconsistent data access destroys confidence in AI, what fragmented architecture actually looks like, and how organizations can build the foundation that makes AI outputs reliable at scale.

https://vist.ly/54v56

Most enterprise AI failures are actually data access failures. Learn why fragmented integrations, inconsistent governance, and disconnected systems prevent AI initiatives from scaling — and how unified data access architectures and MCP are changing the equation.

05/19/2026

A lot of organizations are still thinking about AI as chatbots and meeting summaries.

Microsoft is already thinking several steps ahead.

This carousel highlights key questions around Microsoft 365 E7 and Agent 365.

If you want a clearer picture of how Microsoft is approaching AI governance, identity, and operational AI, check out Korey Stillman’s full article:

From Copilot to Agent 365: What Changes in the Microsoft E7 AI Stack

Link in comments.

05/18/2026

In traditional environments, identity management was mostly about people: authentication, SSO, MFA, and role-based access.

In AI-enabled environments, identity becomes the control plane for the entire operation. Organizations aren't just managing employee access anymore; they're managing trust relationships between users, devices, applications, AI systems, and automated agents.

Korey Stillman's blog on Microsoft 365 E7 walks through why that shift matters, where Microsoft Entra fits, and what organizations should evaluate before moving toward AI-native workflows.

https://vist.ly/54jxy

05/15/2026

When organizations try to operationalize AI across departments, data access becomes the bottleneck.

In this blog, Aaron Varga walks through why. One team builds a Salesforce integration. Another builds a slightly different one months later. A third builds another for a different agent platform. Before long, you're maintaining dozens of fragmented connectors, each with its own access controls and logging.

MCP (Model Context Protocol) provides AI systems with a standardized way to access enterprise systems, so you can build reusable access layers once, rather than rebuilding integration logic each time.

Click here to read his full blog: https://vist.ly/54brj.