FX Repatriation for Lean Middle-Office Desks

About This Project

Middle office analysts running FX repatriation work as the integration layer between systems that don't communicate — reconciling prime broker files against internal books, timing conversion against rates that may already be stale, documenting every step under regulatory pressure. I designed the desk tool for the lean middle office at small hedge funds and fintech treasuries. Fund-strategy rules trigger each recommendation, the AI explains its reasoning, and every action logs to an audit trail. The operator decides.

OmniRisk is one workflow within a broader middle office operations platform. This case study scopes deliberately to the FX Repatriation workflow within the FX Management section — the most thesis-anchoring example of the platform's design approach.

The goal was not to build another data terminal. It was to design a tool that earns trust by making every AI recommendation explainable, every workflow auditable, and every action recoverable.

The Problem

Three things make this hard to solve well:

∙ The workflow has no home. Every existing tool owns one slice — rates, allocations, execution — but none sits inside the repatriation cycle itself. The analyst is the connective tissue, and that's not a UI problem.

∙ The users are sophisticated skeptics. They've spent years building workarounds that function, and they know exactly where the workflow breaks. Anything that can't explain itself, audit itself, and defer to their judgment gets rejected.

∙ AI makes it harder before it makes it easier. The features most likely to help — automated reconciliation, rate forecasting, AI-timed execution — are the ones most likely to trigger resistance without transparency. The question isn't whether to include AI; it's how to earn the right to.

Why This Project

I spent over fifteen years in institutional finance operations across AllianceBernstein, Mizuho Alternative Investments, and Bloomberg LP. FX repatriation was a recurring weekly responsibility. I lived inside the tooling gaps this case study describes — the disconnected systems, the manual reconciliation under time pressure, the trade decisions made without forecast data, the audit trail entries written between phone calls.

The decision to design for this workflow rather than around it reflects that operator experience. The trust thesis — that AI in this space must be explainable, controllable, and auditable as conditions for adoption — is not a research finding I discovered. It is a condition I would have demanded as the operator. The research confirmed it across four other professionals in adjacent roles.

This case study is what happens when someone who ran the workflow operationally for fifteen years designs the tool they wished they had.