Manifesto.
A statement on what we believe, what we refuse to do, and why we think the next twenty years of finance gets rebuilt - starting now.
Finance is the operating system of the modern company. When it runs well, no one notices. When it runs late, every decision in the company runs blind.
Most finance teams today run on infrastructure built before the iPhone. SAP shipped in 1972. Oracle Financials in 1987. NetSuite in 1998. Each was a generational leap when it arrived. None has been fundamentally rebuilt since. The teams using them have invented spreadsheets, scripts, and stitches to bridge what the systems can't do - and somewhere in there, a profession that should be defining the future of business has been reduced to closing books.
We started Rever because we ran finance functions ourselves and we wanted something different. Not a faster ERP. Not a smarter dashboard. A system that thinks first - about the numbers, about the decisions, about the work the people running finance should actually be doing.
What follows is a statement of what we believe and what we refuse to do. We are publishing it now, while the company is small, so that it can be held against us as we grow.
Four observations worth a company.
Books close in cycles when data lands in real time.
The monthly close exists because reconciliation used to be manual. It hasn't been for a decade. The cycle persists because the systems assume it. Closing in cycles made sense in 1998. It hasn't made sense since the cloud arrived.
Operators decide blind.
Every contract, hire, vendor renewal, and pricing change moves the P&L. The person making that decision rarely sees the financial impact at the moment they decide. Finance sees it weeks later, with no link back to the decision that caused it. Variance becomes a forensic exercise, not a steering input.
Data fragments by architecture, not by accident.
Every system in the finance stack was designed to be the source of truth for itself. Combine them and you get six sources, three spreadsheets, and a controller stitching by hand on Monday morning. The standard fix is an "integration project" that takes a quarter, requires data engineers you don't have, and breaks the moment a vendor changes an API.
“Anomaly detection” is mostly rules engines.
Most finance software ships with threshold-based rules. They catch what you predicted and miss what you didn't. The errors that matter - pattern breaks, fuzzy duplicates, policy violations that technically passed - slip through the rules and reach the books. The auditor finds them eight months later. By then, three quarters of decisions have been made on numbers that were quietly wrong.
Seven convictions that shape the product.
Finance is infrastructure.
Not a back-office function. Not a reporting layer. The system through which every operating decision becomes visible - or doesn't. Companies that treat finance as infrastructure run differently from companies that treat it as overhead. We are building for the first kind.
The right person should know the financial impact at the moment they decide.
Not the finance team, weeks later. The operator. With the math attached, the alternatives modeled, and the routing handled. Decisions made with the number in front of them are different decisions. That's the wedge. That's why this matters.
Books should close themselves.
Continuous reconciliation, parallel verification, automated accruals. The close becomes review, not reconstruction. The 200-person company that spends three FTEs on close every month should be spending zero. That work doesn't move the business - it resolves data the system should have resolved at ingest.
Audit-grade by construction, not by retrofit.
Immutable records. Bitemporal history. Source attribution on every cell. Auditors get evidence pre-formatted. SOX walkthroughs become queries, not projects. Compliance is a property of the architecture, not a layer you bolt on after the books close.
Architecture eats features.
A new SaaS feature can be built in a quarter. A new architecture cannot be retrofitted onto a twenty-year-old data model. Most "AI-native" claims in our category are LLM features bolted onto pre-cloud foundations. The system that gets the foundation right today will run finance for the next twenty years. We intend to be that system.
Software should not own your data.
Read-only by default. Self-hosted on request. Your tenancy, your residency, your perimeter - your decision. We are willing to lose deals to remain on the right side of this. We have already turned down customers who asked us to relax it.
Regulated industries deserve infrastructure-grade software.
Defense. Finance. Healthcare. Energy. The teams operating in these industries cannot use cloud-only SaaS. Most modern AI tools are built for them last, if at all. We were built for them from the foundation up - and the architecture decisions that serve them serve every other customer too.
Lines we won't cross.
Every company eventually has to choose what it won't do. These are ours, written down while we are small enough to be held to them.
Three forces have converged.
The technology arrived in 2024. LLMs cleared the accuracy threshold for finance work - reconciliation, categorization, and anomaly detection now run at over 95% without hand-built rules. Until eighteen months ago, this was infrastructure-level impossible. The systems we are building today could not have shipped in 2022.
The incumbents cannot follow. NetSuite is twenty-six years old. SAP is over fifty. They can ship AI features in a quarter; they cannot rebuild their architecture in a decade. Rebuilding the foundation is a once-a-generation event - and the system that gets it right now runs finance for the next twenty years, the way NetSuite has since 1999.
The buyer is pulling. Forty-six percent of accountants now use AI daily. Finance teams are short-staffed and ready to switch. CFOs are not being sold to anymore - they are hunting. The category will be defined in the next eighteen to twenty-four months. We intend to define it.
A system that thinks first.
We are building what we wished existed when we ran finance functions ourselves.
Not a better dashboard. Not another reporting layer. A system that thinks first - about the numbers, about the decisions, about the work the people running finance should actually be doing.
If that sounds like the company you want to build with, you'll know within ten minutes of seeing the product. If it doesn't, we wish you well - and we'll see you in eighteen months.
Founders, Rever
Ready to see it?
Start with the product. Talk to the founders. Read the architecture deep-dive. Whichever entry point matches how you work - we'll meet you there.