In October I accepted a new role as a Product Manager for SAS Viya Model Studio. My new responsibilities mean in addition to knowing how the platform works, I also have to know why it exists, who it serves, and where it fits in a crowded analytics landscape.

This series documents that journey of understanding. It's equal parts learning expedition and product evangelism. A way for me to get smarter about the platform I own while making it more accessible to data scientists and analysts who might benefit from it.

Why An Analytics Platform

When I was a data scientist I loved using open-source tools like Python and scikit-learn. And as my skills developed I branched out into other areas like MLOps. Before leaving for this new role I was on teams building an internal Generative AI tool and predicting customer churn. It was interesting, challenging, and rewarding.

Like all development teams we ran into issues. Ensuring our modules could work together in the intended order and that our data was always properly formatted was always a struggle. I learned a lot building and maintaining our analytic applications. But I wasn't very efficient.

It's important to keep in mind that my core responsibility was making reliable predictions. The faster I could do that the better I was at my job. Building and maintaining our code bases was fun came with a lot of the job training that delayed results. Having access to a platform that could develop useful patterns that could be reused for similar models, build reports that explained results, or deploy a models at scale would have been helpful.

Enterprise platforms promise to solve these issues. But they often introduce new ones. Many are black boxes that hide crucial implementation details. Others are so configuration-heavy that simple tasks become burdens. The best platfoms strike a balance between handling operational complexity while preserving transparency and control.

That's the promise of SAS Viya Model Studio.

What You'll Learn

We'll explore Model Studio starting with environment setup and build toward production deployment. Here's my plan:

Posts 1-3: Foundation and Environment Setup — Getting oriented, setting up your workspace, and understanding the Model Studio architecture

Posts 4-7: Core Workflow Capabilities — Building pipelines, exploring data preparation, training models, and comparing results

Posts 8-10: Advanced Features and Integrations — Model interpretability, Python/R integration, APIs, and automation

Posts 11-12: Production Considerations and Lessons Learned — Deployment patterns, monitoring, governance, and a retrospective on what I wish I'd known

Next Up

In the next post, we'll tackle how you can get access to Viya and understanding the architecture.

Please reach our if you have any questions or recommendations.