What is Knemora?
Knemora is an open-source, local-first knowledge workspace for people and organizations working with private documents and datasets. It compiles source material into a searchable Markdown Wiki and connected knowledge graph, then uses local models to answer questions with citations. Ordinary document processing, workspace storage, and inference stay on your workstation, while optional public search is a separate, off-by-default path. The project is in alpha; automatic setup is currently verified for macOS on Apple Silicon.
Knemora Features:
Document Intake: Import PDF, TXT, Markdown, DOCX, CSV, and XLSX files, then follow their validation, extraction, analysis, and Wiki compilation stages.
Source Provenance: Preserve original files and attach claim-level references, including source identity and locators, to the knowledge derived from them.
Markdown Wiki: Compile approved knowledge into readable Markdown pages with explicit links that remain inspectable outside the application.
Knowledge Graph: Browse explicit, evidence-backed relationships between Wiki pages through a native graph, with lexical search and bounded graph navigation.
Cited Local Chat: Ask questions about authorized knowledge in Wiki mode and receive grounded answers that link back to their supporting evidence.
Dataset Analysis: Analyze CSV and XLSX data with allow-listed operations for filtering, aggregation, statistics, and chart-data generation.
Knemora Benefits:
Privacy Boundary: Keeping routine processing and model inference on your workstation reduces the need to send private workspace material to an external provider.
Evidence Review: Source references let you check where a claim came from instead of relying on an answer without its supporting material.
Portable Knowledge: Readable pages, preserved originals, SQLite metadata, and portable backups make the workspace easier to inspect and move.
More Consistent Analysis: Deterministic operations on structured data help avoid relying on a language model to perform calculations.
Prepared Offline Work: After initial setup, private work can continue without a hosted inference service or an internet connection.
Use Cases:
Policy Research: Compile internal policies and review documents into linked pages, then ask questions that point back to the relevant source passages.
Project Briefing: Bring project documents together so a team can browse related requirements and check answers against their original files.
Research Archives: Organize a collection of reports and notes into a searchable knowledge base while retaining the source files and their provenance.
Budget Review: Use CSV or XLSX analysis to filter equipment costs, calculate aggregates, or prepare chart data alongside related project documents.
Sensitive Work: Prepare a private workspace for offline use when documents or datasets should remain on a trusted local machine.

