Back to directory

Reference Implementation for A2A-Compliant Pro-Code Agents on SAP BTP with Joule Integration and RPT-1

A modular reference implementation for building enterprise agents on SAP Business Technology Platform that integrate with SAP's Joule assistant over the A2A protocol and use the SAP-RPT-1 model as a forecasting tool.

Reference Implementation for A2A-Compliant Pro-Code Agents on SAP BTP with Joule Integration and RPT-1

This repository is a reference implementation for building A2A-protocol-compliant agents on SAP Business Technology Platform (BTP), showing how a third-party agent can integrate with SAP's Joule assistant, which acts as the client that talks to BTP-hosted or remotely hosted agents over the Agent2Agent (A2A) protocol.

It is built with LangGraph (TypeScript) on SAP's Cloud Application Programming model, integrated with the SAP Cloud SDK for AI to reach models through the Generative AI Hub, and includes human-in-the-loop patterns. The example uses SAP-RPT-1 as a tool to run forecasting use cases, such as predicting payment delays or modeling a new product introduction.

Asynchronous communication is implemented with A2A push notifications over webhooks, including a WebSocket-based server and a React UI that visualizes status updates as they arrive. The project also documents infrastructure-as-code for deployment and includes examples of custom logic extensions.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.