The problem with ledgers
Ledgers are the beating heart of financial systems. From upstart fintechs to global enterprises, organizations spend significant engineering time reinventing ledgers to power their products. Accounting correctness is nearly impossible to achieve without these structured journals of accounts, transactions, and balances. But bearing the responsibility and ongoing cost of building, operating, and future-proofing this infrastructure is wasteful. In aggregate, the scale of engineering effort and time duplicated across organizations and industries is enormous.
Ledger technology should be leveraged, not reinvented.
Over the course of my career as a financial systems engineer, I've been tasked with building infrastructure at nearly every level of the modern financial stack: from greenfield banking cores to direct integration with Visa and other card networks, and even launching the first full suite banking-as-a-service platform in the U.S. on top of BBVA's core. For over a decade, our team has built, operated, and scaled dozens of ledger systems. Looking back one thing is certain: ledgers are classic undifferentiated heavy lifting.
Our team has now met with hundreds of engineers responsible for homegrown ledgers and observed some very common anti-patterns. First, product features often require knock-on code changes to core ledger systems to solve a problem that is orthogonal to the main purpose of the ledger. Or worse, there is no architectural distinction between product and ledger. This lack of a strong product-ledger boundary manifests as brittle applications and muddled data models. Conversely, without the right ledger abstractions, product engineers are burdened with understanding the accounting implications behind every transaction that occurs. When normal product iteration drives a continual cycle of ledger-related refactoring or rewrites, product velocity suffers.
This challenge scales up as organizations grow. Conway's Law suggests that homegrown ledger systems will begin to mirror organizational structure. Facing an array of internal ledgers, data and finance teams have the daunting task of reconciling inconsistent, ad hoc accounting models, which amounts to a financial Rube Goldberg machine that is brittle, expensive to change, and painstaking to audit.
When we look at the alternatives to these homegrown systems, it's clear to me that ledgers are a missing building block in the cloud. We see core banking services that cost millions in licensing fees and are designed to support traditional banking products, or vendor-bundled ledgers tacked on to an otherwise unrelated financial services roadmap.
Personally, these are not the tools I want to use, and I think engineering teams deserve better. That's why I founded Twisp.
Meet Twisp: a modern, cloud-based ledger
At Twisp, we believe ledgers must become a first-class, cloud-native service, designed to fit in seamlessly with your cloud stack. Instead of engineers reinventing accounting from first principles, ledgers need to have foundational accounting primitives baked in. Double-entry accounting should not be something application services need to orchestrate, but rather a simple configuration detail. Instead of endless refactoring and coordination, core accounting must support rapid product iteration.
Our aim is to provide ledger infrastructure with the operational and scaling guarantees of DynamoDB, the correctness guarantees of a formally verified transaction model, and all of the accounting primitives needed to build and operate mission-critical financial products at any scale.
In spirit, ledger systems should look and feel like database technology, not a fintech API.
To fulfill that vision a modern cloud ledger must:
- Operate like cloud infrastructure, not SaaS
- Provide full autonomy and control of financial data
- Be composable for any use case
Twisp provides a framework for engineering teams to define their own custom accounting logic, ledger workflows, and complex balance calculations. As accounting structure is encoded and revised over time, or new use cases added, those changes are automatically provisioned into APIs controlled and configured by the customer. This allows Twisp to be used for complex ledger applications and provides for full control over product life cycles.
We're here to help engineers and product teams own and control their ledger data, without the need to build from scratch or depend on external vendor roadmaps for their own customer data.
What a modern ledger unlocks
With Twisp, in a matter of minutes you can now deploy a ledger and accounting system that scales from zero to hundreds of thousands of complex TPS, all with best-in-class developer experience and cloud ergonomics.
Twisp acts as an immutable source-of-truth or system-of-record for any financial transaction, providing:
- Cloud-native ledgers with fine-grained security controls, horizontal scaling, multi-tenancy support, strongly consistent balance calculations, and rigorous correctness guarantees.
- Accounting templates that provide a common set of tools for card processing, money movement, FBOs, foreign exchange, brokerage accounting, and more.
- A financial workflow engine to build unique funds flows, processor integrations, direct-to-bank treasury services, cash and data reconciliation processes, and any other orchestration procedure.
Lastly, unlike alternatives that price on customer accounts or payment volume—effectively a revenue share—Twisp is priced as a technology solution based on reads, writes and storage, where costs scale to zero for inactive accounts.
Twisp is live and helping organizations rethink their ledger systems. We're working with banks and neobanks, issuer processors, brokerage firms, cash management systems, payment facilitators and more. If you're building, rebuilding, or rethinking your ledger and accounting systems, we'd love to talk. Sign up below to request access or get a demo.