SAS debuts custom AI models

In the latest wave of its $1 billion investment in industry solutions, data and AI leader SAS has debuted a new selection of AI models. Each model addresses a specific labour and time-intensive process that can drag business down.

SAS’ packaged models come either ready-to-go or intended to tailor and accelerate model training on customer data. All can quickly and easily integrate with the existing systems of organisations of all sizes.

Inspired by guidelines from SAS' Data Ethics Practice, customers can rely on SAS' simple documentation to create models that deliver results that are easy to understand and explain. Users are equipped to take back their time and enjoy boosted productivity and return on investment, all while adhering to the highest standards of responsible innovation.

The models, available now, include:

• Cross-Industry:

• AI-Driven Entity Resolution.

• Document Analysis.

• Health Care:

• Medication Adherence Risk.

• Manufacturing:

• Strategic Supply Chain Optimisation.

• Public Sector:

• Payment Integrity for Food Assistance.

• Tax Compliance for Sales Tax.

“SAS Models are based on SAS’ core assets, talent and intellectual property from its wealth of experience working with customers to solve industry problems,” said Kathy Lange, Research Director at IDC. “Between its scalability and seamless integration with existing environments, SAS Models are a great option for those looking to accelerate time to production who might lack the expertise or time to build models from scratch.”

A model portfolio on a roll and an agentic sneak preview

As SAS continues to lead the future of AI for business, new models will be launched per evolving customer and market demand.

Fresh offerings coming in 2025 will include:

• Banking:

• Fraud Decisioning for Payments and Card Models.

• Health Care:

• Payment Integrity for HealthCare.

• Manufacturing:

• Worker Safety Monitoring.

• Public Sector:

• Tax Compliance for Individual Income Tax.

As SAS redefines what’s possible with pre-built AI innovation, the next chapter includes agentic counterparts to SAS’ models, bringing even more autonomy to industry-specific offerings.

It can take weeks or months for data scientists to create and fine-tune data lakes: essential repositories that can store, then crunch large amounts of raw customer data in its original format, then restructure and transform it to make it ready for models to use.

SAS’ answer to this issue: a pre-built AI agent that will automate complex data preparation tasks, helping models run in real time, with no manual data restructuring required.

“We believe the future of AI lies in agents that are not only intelligent but also responsible, ready-to-use and relevant,” said Udo Sglavo, VP of Applied AI and Modeling, R&D at SAS. “Our new industry-specific models, built on decades of domain expertise and guided by our ethics-first approach, represent a bold step toward agentic AI: solutions that think with context, act with purpose and deliver real-world impact.”