Datarails, an FP&A solutions provider, declared FP&A software is dead and AI killed it. The company paired this assessment with the announcement of a new AI-driven platform, FinanceOS.
The new platform enables companies to consolidate and collect all of their financial and operational data in one place, including audit trails, permissions and governance, and use that data to have a variety of different AI models execute financial workflows such as drafting financial reports, generating board-ready presentations, deploying AR agents, running live forecasts and scenario analysis, automating month-end close processes, and orchestrating complex finance operations with accuracy and control.
"You no longer need traditional FP&A tools to build models or run analysis. AI engines like Claude in Excel can generate sophisticated financial models in seconds," said Didi Gurfinkel, CEO and co-founder of Datarails, in a statement. "But intelligence is no longer the limit — infrastructure is. Without a governed operating layer, those models cannot run on real-time, accurate and fully auditable data in a secure, controlled environment."
Jonathan Marciano, vice president of brand and communications at Datarails, said in an email that FinanceOS isn't FP&A software in the traditional sense, as it doesn't prescribe how users build models or run analysis. Instead it provides the governed data infrastructure that lets AI do that work, on live, consolidated, fully auditable data. While the OS portion of the product name is more of a metaphor than a specific software description, it was an intention: Just as an operating system doesn't do your work for you but creates the environment in which applications can run reliably, FinanceOS feeds the AI engines with the user's real data to execute financial workflows. And so while financial planning and analysis still happen on top of FinanceOS, the software itself isn't doing it anymore; financial teams' AI engines of choice are.
This includes Claude, ChatGPT, Microsoft Copilot, Gamma, Lovable and Base44, among others. Further, finance teams can connect to over 400 data sources — including ERP, CRM, HRIS, payroll and billing systems, from NetSuite, SAP, and Sage to Salesforce, HubSpot, BambooHR and industry-specific specialist platforms. Marciano said the company considers FinanceOS not as a replacement for the current set of FP&A solutions currently offered but instead as a new category. It serves a different need than the company's other offerings, for teams to be able to build their own AI tools and workflows with auditability, traceability and confidence in the numbers. He said FinanceOS is the foundation of their existing products, and now is available directly to customers who want that infrastructure layer without the full managed suite.
"In the AI era, we quickly realized that the infrastructure itself is increasingly the most valuable part and invested heavily in this infrastructure and consolidation of financial data for the CFO's office," he added. "Now AI tools have opened up unlimited possibilities for what customers need to build on top of it — with the assurance of a governable data layer. … The existing product suite is designed for teams that want Datarails to handle the workflows end-to-end. FinanceOS is for teams that want to use Claude, ChatGPT, Copilot or other AI tools directly, and need a governed data layer to make that work reliably. It opens up the platform to a broader market — including mid-market companies that weren't ready to adopt a full FP&A solution but are absolutely ready to start using AI for finance,"
Datarails has chosen to go with a usage-based pricing model versus the more common subscription-based approach. This choice was based on the idea that pricing should be connected to value delivered, not how many people use the software. Marciano noted that the SaaS per-seat model was built around the assumption that humans are the primary users of software. As AI agents increasingly do the work that people used to do, that model breaks down. There are fewer human users but more value is generated, and the spokesperson said per-seat pricing doesn't reflect that. Further, usage-based pricing increases accessibility, as people are not necessarily bound into a large seat-based contract.
There's also a practical aspect, Marciano added, as FinanceOS is meant to become operational within a few business days. A usage-based model is easier and faster to start up, as opposed to negotiating annually on a per-seat price.
For March, there us a $500 "temporary entry point" offered as part of the early rollout, which includes both the core platform and an estimated level of usage that covers what Datarails estimates a typical mid sized finance team would need.
"Every other solution in the market either locks you into their proprietary AI and workflow stack, or leaves you on your own to figure out the data infrastructure," said Marciano.
