TraceMatter

TraceMatter is a local-first workspace for organizing, validating, and analyzing complex collections of PDF documents.

It is designed for people who need to understand large case files, administrative records, legal documents, or other evidence-heavy material without losing the connection between every finding and its original source. TraceMatter aims to transform scattered files into a structured, searchable, and verifiable record while preserving document identity, page references, chronology, and provenance.

The current prototype implements a tested end-to-end workflow using synthetic data. It runs locally without API keys or network access. The project uses Codex and GPT-5.6 as development partners for architecture, implementation, testing, and documentation; the application itself remains designed around transparent and reproducible processing.

The long-term goal is to help users move from document overload to a reliable evidence map: what happened, when it happened, who was involved, which source supports each statement, where contradictions exist, and what still requires review.

Built With

• codex
• gpt-5.6
• pytest
• python
• streamlit