DatumInt is a lightweight data inspection tool for catching file-level data issues before they silently break pipelines, dashboards, or downstream systems.

In real-world data workflows, many problems don’t show up as hard errors. Files parse correctly, schemas technically match - yet issues like missing fields, inconsistent types, placeholder values, encoding quirks, or subtle structural drift still cause confusion later.

DatumInt focuses on this early boundary: inspecting CSV, JSON, and similar files to surface problems that usually slip past basic validation. It’s designed for engineers, analysts, and builders who want fast visibility into messy or unfamiliar data without setting up heavy contracts or writing custom scripts.

This is an early-stage project, currently optimized for small files and exploratory use. The goal is to better understand which file-level data issues are most painful in practice and build toward clearer, more reliable ingestion workflows.

If you work with real data and have ever thought “this file looked fine, but something was still wrong”, I’d love your feedback.