What Is Scholaris?

Scholaris is an AI-powered research assistant that transforms the way you interact with academic papers, reports, and other documents. Instead of manually skimming through dozens of PDFs to find a specific passage or figure, Scholaris lets you ask natural-language questions and get precise answers drawn directly from your personal library.

At its core, Scholaris converts your documents into a rich semantic format called SPDF, which enables deep understanding of every paragraph, figure, and table in your files. Once converted, your documents become fully searchable using state-of-the-art AI models that understand meaning rather than just matching keywords.

Everything runs 100% locally on your own hardware. Your documents never leave your machine, which means your research stays private and secure by default. There are no cloud subscriptions, no data sharing, and no usage limits beyond your own hardware capacity.

Installing and Setting Up Scholaris

Getting started takes just a few steps. Scholaris consists of a web-based frontend and a local backend that handles AI inference.

System Requirements

Before you begin, make sure your system meets these minimum requirements:

Operating system: Linux, macOS, or Windows (WSL2 recommended)
RAM: 16 GB or more
Storage: At least 10 GB free for models and documents
GPU (optional but recommended): An NVIDIA GPU with 6+ GB VRAM dramatically speeds up document processing

If you have an NVIDIA RTX 3060 or better, you can expect a 15-page PDF to be fully processed in about three minutes. CPU-only mode works too, but processing times will be longer.

Running the Application

Start the backend server and the frontend in separate terminals:

# Terminal 1 - Start the backend
cd backend
source .venv/bin/activate
uvicorn app.main:app --host 127.0.0.1 --port 8765 --reload

# Terminal 2 - Start the frontend
cd frontend
npm run dev

Once both services are running, open your browser and navigate to http://localhost:3000. You should see the Scholaris dashboard ready for use.

Uploading Your First Document

With Scholaris running, adding documents to your library is straightforward:

Navigate to the Documents page from the sidebar.
Click the Upload button and select a PDF from your file system.
Scholaris will begin converting the PDF into SPDF format. You can monitor progress in real time through the conversion status indicator.
Once processing is complete, the document appears in your library with full text, metadata, and embeddings ready for search.

Scholaris extracts text using advanced OCR models, so even scanned PDFs and image-heavy documents are fully searchable after conversion. Metadata such as title, authors, and publication date is automatically detected where available.

Searching Across Your Library

This is where Scholaris truly stands out. Once you have documents in your library, you can search across all of them using natural language.

Type a question like "What are the main limitations of transformer architectures?" into the search bar. Scholaris uses semantic search to find the most relevant passages across every document in your library, ranked by meaning rather than simple keyword overlap.

Search supports several powerful modes:

Semantic search finds passages by meaning, even if they use completely different wording than your query.
Hybrid search combines keyword matching with semantic understanding for the best of both worlds.
Cross-modal search can match text queries against figures, diagrams, and video frames in supported document types.
Intelligent search uses a local LLM to expand your query, detect intent, and provide answer extraction directly from the results.

Results include precise citations with page numbers, so you can always trace an answer back to its source.

Using the Auto-Citation Feature

Scholaris includes an auto-citation tool that helps you generate properly formatted citations for any document in your library. When viewing search results or browsing a document, you can generate citations in common formats such as APA, MLA, Chicago, and BibTeX with a single click.

The autociter extracts metadata directly from the document's SPDF representation, including author names, publication dates, journal titles, and DOIs. You can also export citations in bulk using the BibTeX export endpoint, which is useful when assembling a reference list for a paper or thesis.

Next Steps

Now that you have Scholaris up and running, here are a few things to try:

Create libraries to organize documents by project, course, or topic.
Explore cross-modal search by uploading documents with figures and querying for visual content.
Read about the SPDF format to understand what makes Scholaris search so powerful: What Is SPDF?
Try different search modes to see which works best for your research workflow.

Scholaris is designed to grow with your research. The more documents you add, the more powerful your personal knowledge base becomes, and because everything stays on your machine, your data remains yours alone.