Sherpa of Data

Category: AI

  • Governance Without Slowing Everyone Down

    —

    by

    Kevin
    in

    At this point in the series, we’ve talked a lot about visibility, exposure, and risk. And that’s necessary. But eventually, every governance conversation runs into the same wall: “This sounds great… but people still need to get their jobs done.” That tension is real. Because the fastest way to make governance unpopular is to make…

  • Using AI to Improve Access Governance Instead of Making It Worse

    —

    by

    Kevin
    in

    So far in this series, AI has mostly been the thing exposing the problem. And fairly so. AI amplifies access models.It traverses systems quickly.It exposes weak governance faster than most organizations are prepared for. But here’s the part that gets overlooked: AI can also become one of the most effective tools for understanding access complexity.…

  • How to Clean Up Access Control Without Breaking Everything

    —

    by

    Kevin
    in

    At some point, every organization reaches the same conclusion: “We need to clean this mess up.” Usually after: That realization matters. But it’s also where many teams make a critical mistake: They treat access cleanup like a technical problem. It isn’t. It’s an operational problem wrapped around a technical system. And if you approach it…

  • Most AI Initiatives Don’t Fail for the Reason You Think

    —

    by

    Kevin
    in

    Over the past few weeks, I’ve been writing about a pattern I’ve seen across a lot of data platforms: AI initiatives start strong… and then stall. Not because of talent.Not because of tooling. Because of architecture. Across different organizations, the pattern is surprisingly consistent: In fact, they’re the default path for many teams. But when…

  • From Legacy to AI -Ready: A Practical Modernization Roadmap

    —

    by

    Kevin
    in

    Part 5 in a series on evolving SQL Server environments into AI-ready architectures. Once organizations recognize the architectural gap between traditional platforms and AI workloads, and understand why lift-and-shift migrations fall short, the next question becomes practical: How do we move forward without disrupting the systems that already work? For most teams, the answer isn’t…

  • Where SQL Server Meets AI: The Case for Hybrid Architecture

    —

    by

    Kevin
    in

    Part 4 in a series on evolving SQL Server environments into AI-ready architectures. Simply moving SQL Server to the cloud isn’t enough. Lift-and-shift migrations reproduce the same bottlenecks — often at higher cost. The friction between traditional platforms and AI workloads persists. The solution isn’t abandoning SQL Server. It’s placing workloads where they belong: a hybrid…

  • The Lift-and-Shift Migration Trap (and Its Hidden Costs)

    —

    by

    Kevin
    in

    Part 3 in a series on evolving SQL Server environments into AI-ready architectures. Once organizations recognize the architecture gap we discussed in a previous post, the natural instinct is often straightforward: “Let’s move the warehouse to the cloud.” Databases migrate. Pipelines move. Infrastructure changes. The environment is now technically modernized – cloud-based, scalable in theory, and…

  • The Architecture Gap Slowing Down Your AI Strategy

    —

    by

    Kevin
    in

    Part 2 in a series on evolving SQL Server environments into AI-ready architectures. In the previous article, we talked about a pattern many organizations are experiencing: AI initiatives begin with excitement… and then quietly stall. Not because the teams are weak.Not because the use cases are flawed. But because the underlying platform was never designed…

  • AI Isn’t the Problem – Your Platform Might Be

    —

    by

    Kevin
    in

    First in a series on evolving SQL Server environments into AI-ready architectures. I see it all the time. Strong SQL Server teams.Stable production systems.Clean governance. Then leadership says, “We need AI.” A pilot launches. A model shows promise. Excitement builds. And then progress slows to a crawl. Not because the team failed. Not because AI…

  • Blueprint for AI in Snowflake: How to Scale Intelligence Without Burning Down the Warehouse

    —

    by

    Kevin
    in

    Because “just add AI” is not a strategy. It’s a dare that can go very wrong. Let’s Recap the Tour Over this series, we covered: Now it’s time to tie it all together into something usable. Not hype.Not vibes.Not “AI transformation journey.” A blueprint. Cue This Must Be the Place (Naive Melody)Because if you’ve done this…