A private, local-first AI assistant that actually knows your files.
I built Document Intelligence because I was tired of "Command+F-ing" through dozens of PDFs and notes just to find one specific detail. This tool transforms your local folders into a searchable, interactive knowledge base using Retrieval-Augmented Generation (RAG).
Unlike most AI tools, this doesn't require you to upload your sensitive data to the cloud. Everything—from indexing to searching—happens on your own hardware.
- 100% Data Privacy: We use local Ollama embeddings (
nomic-embed-text). Your files stay where they belong: on your machine. - Virtual Thread Performance: Built on Java 21, the system uses Virtual Threads (Project Loom) to process and index massive document sets in parallel.
- Verified Citations: No more AI "hallucinations." Every answer includes the specific source filename (e.g.,
(Source: roadmap_2026.pdf)). - Streaming & Thinking: Features a professional "thinking" indicator and human-like streaming response for a smooth, Ollama-style CLI experience.
- Persistent Memory: It remembers your conversation context, so you can ask follow-up questions without repeating yourself.
| Category | File Types |
|---|---|
| Documents | .pdf, .docx |
| Technical | .md, .markdown, .txt |
| Data | .csv, .json |
- Java 21+ (The Virtual Thread magic depends on it).
- Ollama installed and running on
localhost:11434. - Groq API Key: Set as an environment variable:
GROQ_API_KEY.
Open your terminal and grab the embedding model:
ollama pull nomic-embed-text