Skip to main content

InfoConnect Hybrid Search Engine

Turn your documents into a searchable knowledge base using AI.

What you can do:

  • Upload PDFs, Word docs, PowerPoint (PPTX), Excel (XLSX), text files, and images. Legacy PPT/XLS is opt-in (LibreOffice).
  • Search in plain English and get answers with sources
  • Organize documents automatically with categories
  • Use the operator console for search, documents, and cited Chat (Auto / Docs-only / Web)

No programming required — if you can use the command line, you can use the InfoConnect Hybrid Search Engine.

Quick Start (5 minutes)​

New here? Follow these steps to search your first document.

Step 1: Copy env, then start the system​

Copy .env.example to .env before you start services or call the API. The code default for API_KEY_CLIENTS is {} (fail-closed): an empty mapping logs a FAIL-CLOSED banner and every authenticated request returns 401 until you set keys. .env.example includes demo keys such as super-secret-key for local use.

cp .env.example .env
just prod-up
just status

Wait for all green checkmarks. First startup takes 2-3 minutes.

tip

"just" is a command runner — like shortcuts for common tasks. See Justfile Commands for more.

Step 2: Upload Your First Document​

curl -X POST http://localhost:8000/embed \
  -H "X-API-Key: super-secret-key" \
  -F "file=@test-documents/malaysia-wikipedia.txt" \
  -F "document_id=my-first-doc"

Save the job_id from the response.

Step 3: Wait for Processing​

Check until status is completed and search_ready is true:

curl http://localhost:8000/status/YOUR_JOB_ID \
  -H "X-API-Key: super-secret-key"

completed can precede search availability briefly; use the search_ready field from the status response before searching. Takes about 10-30 seconds, plus any snapshot lag.

Step 4: Search Your Document​

Ask a question with an explicit content scope:

curl -X POST http://localhost:8000/search \
  -H "X-API-Key: super-secret-key" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What is the capital?",
    "scope": "content",
    "limit": 5
  }'

You should see Evidence Results. Each row is a Document Version with canonical metadata, lane-local retrieval contributions, and bounded Content and Metadata Evidence Sets. Totals and pagination count Document Versions, not chunks.

What You Can Build​

Enterprise Search — Make company documents searchable
Research Database — Organize papers and articles
Knowledge Base — Build an FAQ from documentation
Document Analysis — Extract insights from contracts and reports

Learn More​

New to the InfoConnect Hybrid Search Engine? Start here:

  1. Installation Guide — Full setup instructions
  2. Your First Document — Detailed walkthrough
  3. Auto-Categorizing Your Documents — Automatic organization

Working with images?

  • OCR Guide — Upload scanned documents and screenshots

Want to go deeper?

API Reference:

Getting Help​

What's Next​

Choose your path:

Just getting started? → Installation Guide

Ready to upload? → Your First Document

Have scanned documents? → OCR Guide