TrustRadius: an HG Insights company

SAP AI Core

Score7.9 out of 10

27 Reviews and Ratings

What is SAP AI Core?

SAP AI Core provides the execution and model-access foundation for SAP’s generative AI hub. It supports enterprise teams building AI applications and agents with SAP and third-party foundation models, centralized orchestration, prompt experimentation, and governance controls.

Read more details.

Top Performing Features

  • Security and compliance

    End to end encryption, GDPR compliance, SSO, role-based permissioning, and other precautions are available to protect proprietary business data.

    Category average: 7.6

  • Managed scaling

    The platform provides the computing resources needed when they’re needed, allowing users to scale training and use up or down.

    Category average: 6.8

  • Model deployment

    After training is complete, models can be integrated into business applications using API endpoints or other convenient, standard methods.

    Category average: 7.1

Areas for Improvement

  • Data monitoring and version control

    Teams can track which data is used in training at which point and roll back to previous versions as needed.

    Category average: 6.8

  • Machine learning frameworks

    A wide variety of machine learning frameworks are available and to be used when training models.

    Category average: 7

  • Automated model training

    After teams begin the process, model training can continue autonomously, enabling faster deployment.

    Category average: 6.7

Pros

  • End-to-end AI/ML model management (development, deployment, monitoring)
  • Seamless integration with SAP systems and Business Technology Platform (BTP)
  • Robust support for the AI lifecycle, including model evaluation and training

Cons

  • Complexity and lack of intuitiveness during initial setup and implementation
  • Difficulties with specific feature functionalities (e.g., SharePoint grounding, PDF generation)
  • Manual validation required for configuration and deployment files

AI core

Use Cases and Deployment Scope

Creation of AI POCs, AI-powered workflows, AI agents - running predictions/ classifications/ summaries on data in HANA Cloud.

Pros

  • Access to models.
  • Orchestration deployment.
  • PII Redaction, translation, guardrails.

Cons

  • Enhanced deployment of our own models. It's very easy to call LLMs, but testing and deploying my own trained models—or a small language model—isn't as intuitive.

Return on Investment

  • Ease of deploying AI POCs - faster return on investment by gaining feedback to finalize the AI workflow we set out.

Usability

Alternatives Considered

Joule

[...] review

Use Cases and Deployment Scope

SAP AI Core is used to automate document extraction process. it is manual and prone to errors

Pros

  • repetitive process

Cons

  • user experience can be improved
  • it still requires lot of coding to enable the models to work

Return on Investment

  • depends on the volume of work

Usability

Other Software Used

Anthropic Claude

SAP AI Core

Use Cases and Deployment Scope

Using the Generative AI Hub features of SAP AI Core for embedded AI scenarios incl. cds oyster custom code scenarios as well as generating content as part of generative AI scenarios.

Pros

  • SAP AI Core provides a great amount of usable models
  • SAP AI Core continuously provides new features

Cons

  • Cost monitoring and token usage
  • Permission and authorisation management

Return on Investment

  • Enable new innovation scenarios for businesses
  • Allow me to choose between various model providers
  • Reduce the number of repetitive tasks

Usability

Other Software Used

SAP Analytics Cloud, SAP Business Technology Platform, SAP HANA Cloud

Go Kangaroo

Use Cases and Deployment Scope

Building CAP based agents,

Pros

  • Environment,
  • Grounding
  • Model selection
  • Model filters

Cons

  • reduce ai unit cost
  • more flexibility in configuration

Return on Investment

  • early to scope it

Usability

Alternatives Considered

SAP Build

SAP Machine Learning for Agent Productivity

Use Cases and Deployment Scope

We have implemented ML analytics for implausible reads which was a great tool which delivered productivity saving and backoffice agent now focus on critical problems

Pros

  • ML for implausible reads
  • Large nos of implausible reads for backoffice to deal with
  • often cause regulatory compliance issue

Cons

  • Better documentation
  • enhancement spots to add custom logic

Return on Investment

  • Productivity saving

Usability

Alternatives Considered

EMMA RPA