From AI Features to an AI Execution Layer
The AI execution layer in enterprise software is the set of agentic capabilities that can read operational data, decide on next steps, and automatically trigger actions across systems without waiting for manual user input. This layer shifts AI from informational tools that summarize and recommend to operational agents that execute business workflows end-to-end. Recent enterprise software acquisitions show this shift in clear terms. Instead of building from scratch, vendors like Asana, Coupa, Salesforce, Vertice, and Aircall are buying platforms that already connect data, workflows, and actions. These deals focus on turning ERP, CRM, spend, and content data into automated work, signalling a move toward agentic ERP systems and CRM workflow automation. The strategic question now is not whether AI can write emails or summaries, but whether it can reliably run business processes inside existing controls and governance.
Asana, Coupa, and Salesforce Target Cross-System Agent Execution
Several recent enterprise software acquisitions highlight how vendors are racing to own AI execution, not just AI features. Asana bought StackAI, a no-code platform for building AI agents that can execute workflows across ERP, CRM, ITSM, document systems, and industry applications, turning Asana’s Work Graph into a hub for human–agent teams that “agentify” complex processes. Coupa acquired Rossum, whose transactional language model powers intelligent document processing for complex invoices, extending procurement automation AI deeper into source-to-pay and tying document intelligence to a large spend dataset. Salesforce, meanwhile, agreed to acquire Contentful so its Agentforce platform gains a native content layer for personalized digital experiences, moving from simple AI recommendations to content-aware actions. Together, these enterprise software acquisitions show a consistent pattern: vendors want AI that connects to real systems, data, and controls so agents can complete work, not only suggest it.
Vertice and Vendr: Building Agentic Procurement Intelligence
In spend and procurement, Vertice’s acquisition of Vendr shows how data scale and agentic workflows are converging. Vertice already processes over USD 75 billion (approx. RM345 billion) in spend and reports a track record of delivering more than 20 per cent savings while doubling procurement cycle speed. Vendr adds trusted benchmarks and market insights from 250,000 negotiated contracts across 32,000 vendors. According to Vertice, the combined dataset will power “purpose-designed AI agents trained on real-world data and tailored to specific procurement use cases.” These agents aim to surface pricing benchmarks, renewal risks, and suggested negotiation tactics directly at the point of decision, inside a single procurement platform. This is procurement automation AI shifting from static dashboards to execution: agents that can draft outreach, recommend terms, and orchestrate approval workflows so finance and procurement teams make faster, more confident purchasing decisions.
Aircall and Piper AI: From Call Logs to CRM Workflow Automation
On the revenue side, Aircall’s acquisition of Piper AI illustrates how the AI execution layer is reshaping sales operations. Aircall already connects voice, SMS, and WhatsApp for more than 23,000 businesses, with AI tools for call summaries and follow-ups. Piper AI adds a revenue intelligence and workflow orchestration layer that captures cross-channel interactions – calls, video meetings, email, messaging, WhatsApp, and field activity – then turns them into structured CRM updates, deal scoring, and automated workflows. Piper customers report cutting CRM data entry time by more than 50 per cent and improving forecast accuracy by 50 per cent. For sales teams, this form of CRM workflow automation means conversations automatically update pipeline health, trigger next steps, and flag deal risk without manual logging. The focus moves from documenting what happened to letting agentic systems decide and initiate what should happen next.
Toward Agentic ERP and CRM Systems
Taken together, these enterprise software acquisitions point toward a new generation of agentic ERP systems and intelligent CRM platforms. Instead of treating AI as a layer that sits on top of data, vendors are embedding agents directly into core workflows in finance, procurement, sales, and content operations. In this model, operational data feeds AI agents that can interpret events, apply business rules, and execute across connected applications. Procurement agents can negotiate or prepare renewals; sales agents can keep pipelines accurate; work management agents can coordinate tasks across tools. The competitive edge is no longer whose dashboard looks better, but whose AI execution layer can safely act on more of the business. As more platforms consolidate AI execution capabilities, buyers will judge ERP and CRM systems by how much real work their agents can complete automatically, not only by how clearly they present information.
