Modernising SAS language applications and why WPS Analytics works so well




20 Sep 2019




By: Scott Thomson, Global Head of Insurance Solutions at World Programming, UK.

“Do I really get value from my SAS language environment?”

Let me try and answer this… there is no doubt that many SAS language users are considering modernising the analytics tooling they use. If you can relate to this there could be various reasons: maybe you have commercial and pricing pressures – you feel you are paying more than you should be, or you don’t want to be tied to a particular software vendor; maybe you have technical drivers – you may be wanting to incorporate open-source technologies, new visual tools, or a move to cloud platforms.

The decision to make a change is not simple. Many businesses depend on analytical tools and processes for mission-critical, line-of-business processing. Down-time comes at huge risk and expense: even staying purely with SAS language and moving from one SAS language compiler vendor to another is not completely risk-free.

Moving to entirely different technologies brings the greatest risks, but there can be great benefits too. You may consider changing or incorporating:

  • Programming languages
  • Visual/graphical data science and analytics tools
  • Compute and storage platforms
  • Private or public cloud infrastructure
  • Managed services or outsourcing

When weighing up the best potential courses of action, there are many factors to keep in mind:

  • Cost and complexity of code and data migration
  • Robustness of open-source package management and support
  • Maintainability and reliability of free software
  • Interoperability
  • Performance and scalability
  • Platform and hosting cost control
  • Data security

Balancing risk, cost and benefit is complex.

The practical difficulties of maintaining 20+ years of SAS language applications and workflows created by business users can be daunting. You know the ones I mean? “multiple extracts from operational systems that get merged, filtered and summarised, passed to the finance team who tweak in large Excel files which end up in complex actuarial calculations for the latest price review session”. If this resonates, the WPS Analytics platform is the answer.

Why listen to me? Since leaving university as an applied statistician in the early 1990s, most of my professional career has involved using SAS language tools in some shape or form – whether working for tier-one insurers, SAS Institute, World Programming or one of the big four management consultancies. My time has been spent variously installing, developing, supporting and presenting on most aspects of SAS language tools ranging from installing SAS System software version 5.18 on mainframes back in 1995, to using the latest versions of WPS Analytics software on desktops and servers, AWS and Azure. I have witnessed the introduction and rise of analytics programming languages, graphical tools, workflows, enterprise data integration, business intelligence suites, web-based data science tools, end-to-end solutions in fraud management, financial risk monitoring, customer marketing, high-performance visualisation and in-memory analytics. I can’t say I’ve seen it all, but I’ve seen quite a lot of it!

So what, I hear you say. I frequently hear people say that they love the power and flexibility of the SAS language but due to rising costs and internal demands to reduce IT spend, they are looking for ways to control costs associated with running their SAS language programs. At the same time there is an interest in moving at least some data and processes from on-premises systems to public cloud-based hosted platforms and incorporating “free” open-source technology.

What can you or should you do? I believe you need three components for a successful and effective data science and analytics environment:

1. One development tool for both visual workflows and coding

Consider a full life-cycle analytics scenario: You want to build a complex analytical process (workflow) using a high-productivity GUI tool using your own logic for data acquisition from Excel, databases, Snowflake and S3, prepare data to shape and cleanse it ready for analysis, develop advanced predictive models and explain those models with decision trees, seamlessly share data between R, Python and SAS code for modelling tasks, publish results to Power BI, Tableau or Qlik and make results available for external processing via APIs. WPS Analytics consolidates these capabilities into a single toolset. With its own fully-integrated SAS language complier, you can combine R and Python with the SAS language and with other commercial and non-commercial tools and technologies to create single executable programs or visual workflows.

With the WPS Analytics approach, it’s simple to see how this differs from the complex approach of needing multiple products offered by other vendors. The open interoperability philosophy of WPS Analytics puts the power of commercial tools together with open source technology at the fingertips of data scientists and business analysts for ease of use and better productivity. All this at a highly-competitive price.

Here is a short video  about WPS Analytics.

2. One tool for deployment, integration and management

Let’s consider “productionisation”: Having developed a valuable analytics program or predictive model, to get real value back from it you need to deploy it to a managed production environment that gets maximum exposure and use out of your predictive model. This should likely involve Restful APIs so that other systems, business users and external services can easily consume the deployed program or model. WPS Analytics Hub supports clear and easy deployment, operation and governance of programs and models into production, with version control integration and clear separation of development, test and production environments.

Watch this informational video about using WPS Hub for Model Deployment.

3. One pricing structure that controls cost whilst you create value

Modern public and private cloud infrastructure enables businesses to scale analytics processes on-demand for development, test and production. This can be a difficult fit with traditional licensing models that typically involve an annual subscription for a fixed size of compute environment or number of “seats” (users). Your data scientists and analysts may, for example, want to work on a large analytical program for just two months. Paying a full annual software license subscription fee just to cover this time-limited project may seem poor value, although this is almost universally still the situation. In addition, increasingly a problem with cloud deployments is internal governance controlling who within an organisation is permitted to leverage a cloud environment. Without a governance programme, businesses risk racking up unexpected cloud platform and/or software licensing bills and falling out of licensing compliance.

For enterprise customers WPS Analytics is available with new flexible enterprise cloud pricing models that are designed to accommodate the issues arising with elastic compute and storage infrastructure. WPS Analytics enterprise cloud pricing works by charging only for production analytics, for data and users. We understand that this may not suit everyone, and our traditional pricing model is still available.

WPS Analytics for enterprise cloud computing is delivered in 3 easy components

  1. Choose an infrastructure cloud (AWS, Google, Azure)
  1. Subscribe to a package for production jobs and transactions
  1. Unlimited analytics for development and testing

Modernising your analytics environment doesn’t have to be a step into the unknown and can bring huge benefits.

Migrating your existing SAS language programs to WPS Analytics is easily achieved in a simple three step approach:

  1. Run the WPS Analytics code analyser on all your existing SAS language programs to generate a detailed compatibility report. Code analysis of thousands of programs takes just seconds to generate a heads-up view on the functional fit of WPS Analytics to your needs.
  2. Deploy WPS Analytics for free and run some of your selected SAS language programs to gain confidence in the software’s capability.
  3. Migrate your applications, programs and data with full support from experts at World Programming if necessary. Migration projects are only signed-off when you are 100% satisfied.

Here is a brief video about Migrating to WPS Analytics

And finally…

Maybe consider the questions in this video,  Yes to WPS Analytics and how they might relate to your current needs and how you plan to future-proof and retain the investment you’ve made in your analytics environment. Don’t underestimate how valuable your data science team and analytics applications are to your business!

For more information, get in touch with our sales team ([email protected]) and let us show you how we can support your investment in analytics or simply sign-up for a free evaluation of WPS Analytics.

Have a question?

Get in touch with our sales team

Try or buy

Standard Edition
Academic Edition
Community Edition