zeb: Putting agentic AI into action as an active participant in consulting workflows

Business needs

Improving project outcomes through AI workflow reinvention

As a consulting firm, zeb understands that successful client projects depend on precise and up-to-date information about project status, milestones, efforts, and dependencies. While business-relevant information is generally consolidated within SAP and accessible via Fiori apps, project managers still need to navigate the correct paths to retrieve the required information, often in a different context. Switching between apps and relying on cross-contextual views does not always meet the need for quick, individualized access to information.

In search of a better solution, the consulting firm decided it was time to explore the use of agent-based AI to transform project information management and increase productivity. The goal was to go beyond purely software-based automation and create an autonomous AI orchestration system that can independently determine the appropriate path to specialized AI agents in order to process specific search and input requests.

Solutions

Scaling agentic AI as a performance enabler for the business

As a critical first step in its exploration of agentic AI, zeb teamed up with SAP platinum partner NTT DATA. With extensive experience designing secure and scalable engineered applications and integrations using services on SAP Business Technology Platform (SAP BTP), NTT DATA was well equipped to lead the way on agentic AI solution development. This included demonstrating the best approach for using this advanced form of artificial intelligence to orchestrate multi-step workflows that interact with SAP systems and data.

NTT DATA then built an orchestrator agent capable of managing and coordinating other agents and used the AI Foundation solution and tapped SAP Cloud SDK for AI to allow access to enterprise data. NTT DATA then developed custom agents with an SAP Fiori UI for deployment on SAP BTP Cloud Foundry environment and connected SAP S/4HANA to SAP BTP using SAP Integration Suite.

These individual agents enable intuitive, natural language-based access to project information, employee assignments, as well as initial predictions and anomaly detection. One agent provides project-related information by leveraging OData services in S/4HANA. Based on Assist-the-AI capabilities, it can deliver not only raw data but also forecasts on budget consumption and highlight potential project anomalies to the user. If additional project staffing is required, another agent can propose and pre-book suitable employees based on availability data and further criteria such as skill profiles.