FXpress · AI Co-Pilot for Cross-Border Remittance
| Property | Value |
|---|---|
| 🏷️ Status | Complete |
| 🎯 Focus | Cross-border remittance · FX timing · AI trust · Stablecoin awareness |
| 🔖 Methods | Competitive analysis · User interviews · Survey · Usability testing |
| 👥 Participants | 4 interviews · cross-corridor · varied crypto comfort |
FXpress is a conceptual companion app — a banking assistant for people who send money across borders, to family and dependents who rely on those transfers to cover rent, groceries, and school fees. The research phase was designed to ground this product in the real conditions under which people make these decisions: under time pressure, with incomplete information, through tools that frequently obscure the costs they're charging.
FXpress occupies a specific position in the financial landscape. Rather than replacing a user's bank, it sits alongside it — a digital-first assistant for globally mobile senders, offering the transparency and timing help that existing remittance tools have historically withheld. What makes FXpress distinct — and its research questions harder — is the integration of two emerging technologies most users approach with skepticism: AI-driven financial assistance and stablecoin-based payment routing.
⚠️ Note: This research supports a conceptual feature set intended for future-state readiness, contingent on regulatory infrastructure and real-world crypto-to-fiat off-ramp maturity. Findings reflect user attitudes and behaviors under current conditions, projected forward.
To design a more seamless and user-trusted experience for converting and sending foreign exchange currencies — domestically and internationally — via mobile. This includes exploring AI-powered FX timing recommendations, smart routing through stablecoins such as USDC, and local fiat delivery via established off-ramp partners.
The research was organized around five core questions, each addressing a different layer of the design problem.
The first objective was to understand how users currently send international payments — what tools they use, what workarounds they have developed, and where their existing workflows consistently break down. The second was to identify the specific pain points around FX timing: what it feels like to not know whether today is a good day to convert, and what the cost of that uncertainty has been in practice.
The third objective addressed the trust dimension directly: what conditions would need to be true for a user to allow an AI assistant to recommend — or partially automate — a financial decision? The fourth explored stablecoin openness, specifically whether users would accept crypto-based routing if the mechanics were invisible and the outcome was denominated in familiar local currency. The fifth examined automation comfort levels: how much of the process each user type wants to hand over, and where the line is between helpful and threatening.