Composable back office: Rippling and AbstractOps

Diving deeper into

Hari Raghavan, CEO of AbstractOps, on the composable enterprise

Interview
A third of our customers are on Rippling, which clearly means that a customer is getting value from both our platforms.
Analyzed 4 sources

The real signal is that Rippling and AbstractOps sit on different systems of record, so they can coexist inside the same customer instead of forcing a rip and replace. Rippling starts from the employee record and automates payroll, onboarding, IT access, cards, and workflows tied to workers. AbstractOps starts from company records like contracts, vendor relationships, payments, and compliance tasks, then connects those tools into one operating layer.

  • The overlap is narrow and mostly at hiring and onboarding. After that, the workflows split. Rippling handles employee lifecycle work, while AbstractOps is built for the corporation level work around documents, counterparties, approvals, and repeatable legal and finance operations.
  • This pattern shows up elsewhere in the stack. Roster uses Rippling, AbstractOps, and Ramp together, and describes each one as organizing a different data model, employee records for Rippling, company records for AbstractOps, and spend transactions for Ramp. That is classic composable software adoption.
  • The strategic prize is still the old ERP bundle. AbstractOps frames the competitive set as SAP and Oracle, not adjacent startups. Rippling has scaled into a $570M annualized revenue platform with more than 20,000 customers by adding modules around its employee graph, which makes it a powerful partner and a likely expanding neighbor.

Going forward, more back office stacks will look like this, a few dense control points linked by APIs instead of one giant suite. Rippling will keep widening from HR into finance and IT, while companies like AbstractOps win by owning the messy workflows that sit between contracts, payments, vendors, and compliance. The winners will be the products that become the default record for one critical layer, then plug into the rest.