At the Agentforce Tour in New York City, Salesforce and MuleSoft unveiled a new vision where APIs, RPA, and AI agents converge as a unified, metadata-driven execution layer, transforming enterprise automation and calling for strategic API productisation and workforce evolution.
At the Agentforce Tour in New York City on December 10, Salesforce and MuleSoft set out a clear reframing of integration: what was once “API plumbing” is being positioned as the execu...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
tion layer for AI-driven automation. According to the original report from ERP.Today, speakers Diane Kesler and Shivani Marrero argued that APIs, robotic process automation (RPA) and AI agents should be treated as a single operating model rather than discrete capabilities, with metadata and event streams supplying the context that allows agents to take reliable action across legacy and cloud systems.
The vendors’ vision places APIs at the heart of enterprise automation. Salesforce’s recent product announcements emphasise APIs not merely as connectivity points but as the structured training data and trusted interfaces that AI agents require to operate safely. Salesforce said in its company release that new AI-powered enhancements to MuleSoft , including Intelligent Document Processing (IDP) and tighter integration of Einstein into tools such as Flow Builder and Anypoint Code Builder , are intended to let business users and developers build integrations and automations with natural language prompts, accelerating time to value.
MuleSoft’s materials reflect the same logic: APIs expose system actions, automation tools assemble those actions into workflows, and AI determines when and how to trigger them. The vendor’s guidance shows how Process and Experience APIs can be adapted into Model Context Protocol (MCP) servers to give agents structured, real‑time access to enterprise data and operations without requiring extensive new architecture. Product capabilities cited by MuleSoft , from Anypoint Platform and API Manager to Composer, Flex Gateway and Agent Fabric , are presented as the governance, observability and execution layers that make agent-driven workflows practical at scale.
RPA is recast in this narrative as a legitimate and persistent complement to APIs rather than a stopgap for poorly integrated estates. ERP.Today reports that presenters framed MuleSoft RPA as an intentional extension point to give agents access to systems that lack modern interfaces, while MuleSoft documentation outlines secure approaches for authenticating and integrating RPA processes , for example, via user API keys and Connected Apps , to maintain governance when bots interact with Salesforce products.
The technical overlay is being accompanied by a cultural and skills reframing. Speakers urged practitioners to evolve from “integration experts” to “automation innovators,” a shift that matches Salesforce’s broader messaging about AI‑powered composability: an AI-centric methodology plus modular building blocks that let developers and business teams compose secure, adaptable AI applications. Salesforce’s public materials position Composer, Flow, IDP, RPA and the Anypoint portfolio as the building blocks of that composability.
For ERP vendors, integrators and in‑house platform teams, the implications are concrete. ERP.Today and vendor briefings together signal rising expectations that core ERP suites expose clean, productised APIs and event frameworks suitable for external orchestration. Industry data shows that enterprise automation programmes succeed when interfaces are governed, observable and decomposed into composable tools that match how agents sequence tasks; organisations that do not prioritise API productisation risk becoming bottlenecks to automation initiatives.
Program and platform leaders will also need to revisit roadmaps and workforce plans. The convergence of integration engineering, workflow design and AI operations suggests new hiring profiles and training paths; partner enablement models may require updates so system integrators can deliver automation outcomes rather than only point integrations. Vendors present RPA as part of a durable hybrid architecture that spans both modern and non‑modern ERP estates, meaning decisions about when to build APIs, when to productise endpoints and when to apply RPA must be strategic rather than ad hoc.
Taken together, the messages from Salesforce, MuleSoft and the Agentforce event describe a future stack that is unified, metadata‑driven and oriented toward safe, observable execution. According to the original report, the immediate ask for ERP insiders is straightforward: expose governed interfaces, embrace composability, and prepare teams to orchestrate AI agents that will rely on those same integration assets to take action. The company announcements and product documentation make clear that vendors are positioning their platforms to be the substrate for that shift , but they also underscore that successful adoption will hinge on disciplined API productisation, security and operations rather than on AI alone.
Source: Noah Wire Services
