The Problem

The Need for a Better Import Flow

In data-heavy financial operations, incoming files from multiple sources often arrive in inconsistent formats, making them difficult to process reliably. These files often lack structure, with inconsistent headers, reporting periods, and naming conventions. As a result, imports would frequently fail without clear feedback, forcing users to rely on guesswork or support.

There was also no safe staging area; even a small formatting issue could overwrite live data. A more reliable, guided import experience was needed to reduce failure rates and give users confidence in the process.

The Solution

A Guided, Error-Tolerant Import Flow

The new workflow breaks down the import process into clear, manageable steps. Each part is designed to prevent errors before they happen and offer safe paths forward when they do.

Structured Error Feedback

Import, Processing & Validation

Once a file is uploaded ected through the import modal, the system begins a structured review in two stages: format checks and data-level validation. Each stage is designed to catch specific issues early and guide the user through resolving them.

Safe Importing – Protecting Live Data

Imported data is kept separate from the live dataset to prevent overwriting or corruption. Users can save drafts and import files even with validation errors, but no changes will impact live data until confirmed by the user, ensuring a safe workflow and maintains data integrity during the import process.

Outcome & Reflection

The redesigned import workflow helped reduce failed uploads and made the entire process more predictable and user-friendly.

• Messy CSVs can now be converted into clean, structured Excel files
• Early validations highlight issues with clear, actionable messages
• An error review modal makes it easy to navigate and fix problems
• Faulty data stays isolated in a staging area until confirmed
• Teams can now complete imports confidently, with less support

Together, these changes improved efficiency, reduced manual back-and-forth, and turned a previously frustrating task into a smoother, more reliable experience.