Automated treasury management: stablecoins, LLMs, and the disintermediation of finance

The world of enterprise software is currently witnessing a breakdown of the status quo. As companies like Craft Docs move away from rigid SaaS platforms (such as Zendesk) in favor of bespoke, AI-powered automation, a fundamental shift is emerging: creation is replacing configuration.

This evolution is no longer limited to marketing or HR. It is rapidly extending into financial infrastructure, specifically targeting the outdated, intermediary-heavy world of cross-border payments and treasury management.

The “Uber-fication” of cognitive tasks

For decades, building custom software was prohibitively expensive for all but the largest organizations. Today, Large Language Models (LLMs) allow even non-technical users to build functional tools through simple prompts.

We are entering a world where software adapts to a company’s workflow in real time. In finance, this shift exposes the inefficiencies of the $7.5 trillion daily FX market still heavily reliant on manual processes and legacy systems.

The death of the correspondent banking intermediary

Traditional treasury management remains slow and operationally intensive. Moving $50 million from USD to MXN often requires multiple manual steps, including checking rates, contacting relationship managers, and waiting days for settlement across correspondent banks.

Programmatic treasury introduces a model where banking becomes largely invisible by removing two key layers of intermediaries:

  • Financial intermediaries: Stablecoins replace slow correspondent banking networks with near-instant settlement.
  • Operational intermediaries: LLMs connected to APIs eliminate the need for manual approvals and communication.

A comparison: legacy vs. programmatic treasury

The legacy workflow (days):

  • Manual entry through terminals
  • Negotiation via calls or emails
  • Settlement delays (T+2 or T+3)
  • Exposure to slippage and operational risk

The programmatic workflow (minutes):

  • API-first systems source optimal pricing instantly
  • Execution occurs through a single confirmation
  • Real-time settlement via stablecoins and RTP rails
  • Humans shift from execution roles to oversight roles

Beyond trading: the future of automation

Programmatic treasury extends across the entire financial stack. In the near future, treasury operations will include:

  • Automated reconciliation across global accounts
  • Real-time visibility of all financial flows
  • Automated hedging based on live market conditions
  • Rule-based investment and yield strategies

OpenFX: the infrastructure layer

OpenFX aims to provide the infrastructure required to support this transition. Similar to how AI platforms offer API-based access to intelligence, OpenFX focuses on enabling programmable cross-border liquidity.

The goal is to support a system where money moves continuously, globally, and at machine speed.

A shift toward programmable finance

As financial systems evolve, competitive advantage will shift away from feature-rich platforms toward flexible, programmable infrastructure.

The next generation of financial leaders will be defined not by their tools, but by how effectively they integrate automation, data, and liquidity into a unified system.