One system of record: how Rippling is building the harness around the LLM
Key points
Key takeaways from Meet Rippling AI, a Kleiner Perkins feature with Rippling (September 2026):
Mission unchanged for a decade. When the team founded Rippling a little over ten years ago the mission was to free smart people to work on hard problems, and the team says it is unchanged today; AI has expanded what is possible for the product and for the engineers building around it.
The data blocker for enterprise AI. Most companies operate tens or even hundreds of software systems whose data does not join cleanly, and Rippling argues that this fragmentation — not the models — has been the fundamental blocker to organisations adopting and maximising the benefits of AI.
One system, one permission model. The Rippling platform is built from the ground up as a single coherent system: one permission system plus a shared set of B2B SaaS building blocks for approvals, reports and workflows that apply across all products.
Data as an unfair advantage in the AI era. Because every product sits on the same data model, Rippling says it can build not just better applications but better AI agents that traverse that data and generate answers that guide how a company is run.
Native AI for every employee, not just admins. Native AI is embedded in the system and available to all employees, so a complicated payroll question — every US state has its own set of deductions — that might take an HR representative hours to answer can be resolved in minutes.
Agents inherit human permissions. Agents act on behalf of somebody and cannot do anything the requester cannot do: an employee cannot ask what their boss's salary is, because neither the person nor the agent holds that permission.
Governing a mixed workforce. One customer came to Rippling to govern its OpenClaw agents — provisioning access and handling onboarding and offboarding — because Rippling was already the best product for the task; the company wants to be the platform firms use to manage the inevitable mix of human and AI workers.
Five-to-ten-year vision: harnesses and orchestration around LLMs. Whether as an HCM, finance or sales system, enterprise software will become specialised domain harnesses with domain-specific tools and work orchestration, with the LLM as the intelligence at the centre.
Read more: Meet Rippling AI on the Kleiner Perkins channel
