The systems of record are not going anywhere. SAP installed itself into the Fortune 500 over thirty years. Salesforce did the same to revenue operations. ServiceNow holds the IT and service-management layer. NetSuite owns the upper end of the mid-market ERP space. These are the systems that contain the actual data of the business. They are slow to install, expensive to migrate, and almost impossible to replace. The a16z piece "Why the World Still Runs on SAP" puts the cost of an ECC-to-S/4HANA migration at $700M, three years, and a fifty-person Accenture team. Lidl scrapped its SAP transition after spending half a billion dollars. Roughly seventy percent of digital transformations fail their objectives, by McKinsey's own estimate.
The records persist. What changes is what sits on top of them.
A new layer is forming above the systems of record
A new layer is forming, and it does not have a settled name yet. a16z calls it the system of action: the layer that reads from systems of record, does work, writes back, and stays accountable for the result. Foundation Capital calls it the system of agents, plural, because the work usually requires several specialised agents collaborating like a team of experts. Bessemer calls it AI-enabled services, distinguishing it from copilots (which augment a human worker) and pure agents (which replace one). The terminology is unsettled. The layer is real, and it is where the next decade of category-defining enterprise software is being built.
What it is not is a copilot. A copilot lives inside an existing application and helps the user of that application do their job faster. The system of action lives between applications and does the job in place of the user. The two are different layers of the stack, addressed at different parts of the customer's budget. Copilots come out of the SaaS budget; systems of action come out of the labour budget, which is roughly an order of magnitude larger.
Why the system-of-action layer is forming now
Why this layer is forming now is a question worth answering directly. The technology to build it has been around in pieces for a long time. Process mining vendors (Celonis, Apromore) have been mapping enterprise workflows since the early 2010s. RPA vendors (UiPath, Automation Anywhere) have been automating those workflows for nearly as long. What neither could do, and what has been the structural bottleneck for a decade, is handle the unstructured-input judgement work that constitutes most of the actual work in a business. Foundation models can handle it. Wornow, Narayan, Ré and colleagues at Stanford put the productivity opportunity at four trillion dollars per year in a 2024 paper that named foundation models as the missing piece process mining and RPA were always waiting for. The pipeline (process mining, then workflow understanding, then execution) is now coherent in a way it has never been.
The vendors of the previous era have noticed. ServiceNow's Zurich release embeds process mining and agentic automation directly into the platform. ServiceNow acquired UltimateSuite (a process-mining vendor) in late 2023 and partnered with Celonis. The category convergence is happening at the platform level. Constellation Research's coverage names the shift directly: you cannot deploy agents reliably without process intelligence underneath them, which is why the formerly separate diagnostic discipline of process mining is being absorbed into the deployment substrate of the new agentic platforms.
What the winning companies will look like
The companies that win the system-of-action layer will not look like the SaaS companies that won the previous layer. They will look like services firms with engineering DNA. Embedded delivery teams in the customer's environment, because the work requires understanding the specific stack and the specific workflows. Outcome-aligned pricing, because the deliverable is measurable and the labour budget is what is being eaten. Vertical depth that compounds over time as the agents trained on one customer's workflows transfer learning to the next. Network effects that come from outcome data aggregated across customers in the same operational pattern.
Bessemer's vertical AI piece argues the TAM for this layer is ten times the combined market cap of the top twenty public vertical SaaS companies. The math is mechanical. Vertical SaaS extracted a margin from a software budget. Vertical AI extracts a margin from a labour budget. The labour budget is an order of magnitude larger.
Why this is not a repeat of the RPA hype cycle
A reasonable counter is that all of this rhymes with the RPA hype cycle of 2018, which plateaued for exactly the reasons most agent skeptics raise now: edge cases multiply, integrations break, governance is hard, the demos work and the production deployments do not. The structural difference is what foundation models can do that RPA never could. RPA could only handle deterministic rule-following on structured inputs, which meant it could only automate the parts of the workflow where the rules were already explicit and the data was already clean. Foundation models can handle unstructured inputs with judgement, which is most of the actual work. The Wornow paper is the formal version of this argument, and it is worth reading in full if you are inclined to file the agent wave next to the RPA wave.
This is why the next decade of category-defining enterprise software companies will not look like the last decade's. The operating system was the platform of the 1980s and 1990s. The system of record was the platform of the 2000s and 2010s. The system of action is the platform of the 2030s. The companies building it now are the ones that will define the next era.
The system of action does not require migration
The system of action does not require the customer to migrate. That is the point of the layer. It runs on top of the existing records, the existing CRM, the existing ERP. The customer keeps everything they have spent thirty years installing. The new layer is what does the work that used to require a person sitting between the systems doing the manual reconciliation, the data entry, the exception handling, the reporting.
a16z's line for what this looks like operationally is the cleanest. The bridge becomes the highway. The integration that used to be a temporary scaffolding between two systems becomes the place where the work actually happens. The system of record stores the truth. The system of action moves it. That is where the next decade is going.