In the study, the team ran through a typical customer churn analysis data and AI life cycle.
See for yourself just how productive the teams were when using SAS Viya compared with the alternatives.
Time to complete the studied data and AI life cycle.
SAS Viya data and AI life cycle is 4.6x faster.
Productivity challenges in the data and AI life cycle:
Complexity, tools and resources
Modern organizations across industries are increasingly data-driven, using data to guide decisions, provide strategic insights, develop new products and drive innovation.
Accelerating the journey from data to decisions enables companies to bring their products to market more quickly, adapt to market conditions with greater agility, reduce operational costs and better respond to customer needs. By capitalizing on data and AI, organizations can ultimately boost competitiveness and drive cost efficiencies.
The end-to-end process of transforming data into decisions is known as the data and AI life cycle.
The life cycle typically involves three core phases – managing data, developing models and deploying insights. These steps form an ongoing, iterative process requiring collaboration among data scientists, data engineers and MLOps engineers. Increasingly, additional stakeholders such as business analysts or ethics specialists may also be required, furthering the need for effective collaboration.
The tasks and team productivity involved throughout the data and AI life cycle are commonly hindered by the time, complexity and resources required.
Each stage of the data and AI life cycle brings about unique challenges, often requiring significant effort, complex tools and teams of expert data professionals.
Manage data
What: Data access, preparation and governance
Who: Data engineer
Develop models
What: Build, optimize and validate AI models
Who: Data scientist
Deploy insights
What: Deploy, monitor and retrain models
Who: MLOps engineer
Overcoming these obstacles can significantly enhance productivity, driving faster decisions, innovation, cost savings, increased revenue and competitive advantages.
- The Futurum Group
To address these challenges, organizations have turned to various approaches, including both commercial and non-commercial data and AI platforms.
While non-commercial tools based on open source technologies can provide significant functionality, they often require considerable expertise to implement and, most importantly, manage.
Commercial platforms typically use the cloud to enhance collaboration and reduce management requirements while providing built-in features and automation to streamline the life cycle.
While there are many data and AI platforms available in the market, their functionality and overall impact on productivity vary.
Achieving 4.6x more productivity with SAS Viya
The Futurum Group conducted a hands-on evaluation of three distinct data environments – SAS Viya, a competitive commercial data and AI platform, and a non-commercial approach where various popular open source technologies were used – to measure the impact that data and AI platforms have on productivity throughout the data and AI life cycle.
Results of the study showed that SAS Viya was found to improve productivity at each stage of the data and AI life cycle. Key to the productivity advantage of SAS Viya was a measurement of engineering time spent completing each task.
The evaluation was achieved by completing an end-to-end customer churn prediction analysis, a common use case relevant to many industries, in each environment. Testing was completed by three Futurum Group analysts designated to specific personas – a data engineer, a data scientist and an MLOps engineer – responsible for completing tasks aligned with the three phases of the data and AI life cycle. Additionally, a fourth analyst evaluated each platform via a business analyst persona to measure if tasks were achievable by a non-technical resource.
Testing showed that an end-to-end data and AI life cycle can be achieved with more than 4x greater productivity in SAS Viya than in competitive solutions. The productivity gains found in SAS Viya not only ease the burden placed on data and AI teams, but they can lead to significant cost savings.
- The Futurum Group
In addition to time savings, The Futurum Group analysts found several key productivity advantages while evaluating SAS Viya compared to the competitive solutions:
Completion of the data and AI life cycle with zero coding required
- Enables participation from users of all technical backgrounds, making the platform suitable for collaboration and enabling a smooth transition between each stage of the life cycle.
- Lowers the learning curve for complex data and AI tasks.
- Eliminates reliance on complex programming libraries.
- While coding is not required, Viya does provide the flexibility to complete tasks with SAS, Python and R code as desired, enabling technical experts to utilize the tools they are most familiar with.
Greater achievability by non-technical resources
- Of the total data and AI life cycle, 86% was found to be achievable by a business analyst – a 30% greater completion rate than the next best alternative approach.
- Utilization of less-experienced resources enables greater staffing flexibility and access to a wider talent pool.
- Unlocks time for expert data practitioners to focus on new innovations, including generative AI-enhanced workflows.
- Greater built-in visualization and automation capabilities enable productivity gains at each stage of the data and AI life cycle.
Lower complexity throughout the entire data and AI life cycle
- Of the tasks completed in SAS Viya, 96% were considered "low complexity," as determined by The Futurum Group analysts. No task was considered "high complexity."
- Only 50% of tasks in the competitive commercial platform and 30% of tasks in the non-commercial platform were considered "low complexity," while 17% were considered "high complexity" in both solutions.
- Lowering complexity enables greater efficiency in completing tasks and lowers the learning curve to onboard new users.
Conclusion
Productivity in data and AI matters
For data-driven organizations, the productivity of their data and analytics teams is vital to core business practices. The tools utilized by these teams can make a significant impact. The current market landscape provides organizations with a variety of approaches, ranging from non-commercial tools to commercial data and AI platforms with various features and capabilities.
This study evaluated the productivity differences in completing the data and AI life cycle across three distinct data and AI platforms and found that SAS Viya presents significant productivity advantages over both commercial platforms and non-commercial solutions.
Overall:
SAS Viya was found to enable 4.25x greater productivity throughout the end-to-end data and AI life cycle compared to the next best solution tested.
SAS Viya was found to be an intuitive platform that significantly lowers the learning curve for new users and enables greater productivity for non-technical users.
These results, combined with the previous Futurum Group testing that found SAS Viya to provide a 30x performance increase over competitive solutions, demonstrate that SAS Viya offers significant overall advantages. The combination of increased productivity and faster computation provides a highly efficient platform capable of unlocking significant value for data-driven organizations.
The selection of an optimal data and AI platform is crucial for enabling organizations to efficiently use the data and AI life cycle to achieve their business goals. This study found SAS Viya to provide a data and AI platform that combines strong technical capabilities with an intuitive interface capable of transforming data into decisions quickly and efficiently.
