In many offices, employees routinely turn to free online tools to convert PDF files, compress images, transcribe audio recordings, or run Optical Character Recognition (OCR) on scanned documents. What many business owners don't realize is that uploading proprietary blueprints, financial audits, or client NDAs to third-party web converters often transfers custody of sensitive corporate data to unknown remote servers.
SnapOtter is an open-source, all-in-one file processing suite designed specifically for on-premises deployment. It provides a clean browser interface, REST APIs, and automated processing pipelines for document conversion, image compression, OCR, and audio transcription—where every byte of data is processed locally on your own server hardware without ever leaving your network.
Core Capabilities of SnapOtter
- PDF & Document Conversion: Convert Word, Excel, PowerPoint, and Markdown files to PDF and vice versa using local LibreOffice/Pandoc engines.
- Local AI Optical Character Recognition (OCR): Extract searchable text from scanned contracts and invoices using local Tesseract/PaddleOCR models.
- Media Compression: Losslessly shrink large high-resolution site photos and video files before archiving.
- Local Audio Transcription: Transcribe voice memos and meeting recordings on your own CPU/GPU using Whisper models.
System Requirements
- A Linux server running Docker and Docker Compose.
- At least 4 CPU cores and 8GB RAM (16GB recommended if running heavy AI transcription pipelines).
- Local storage for temporary processing buffers and converted file outputs.
Step 1: Set Up Directory Structure
Create a dedicated location on your server for SnapOtter configuration and data storage:
mkdir -p /opt/snapotter && cd /opt/snapotter
mkdir -p data uploads exportsStep 2: Create Docker Compose Configuration
Create a docker-compose.yml file with the following configuration:
services:
snapotter:
container_name: snapotter-core
image: ghcr.io/snapotter-hq/snapotter:latest
restart: unless-stopped
ports:
- "127.0.0.1:8088:8088"
environment:
- NODE_ENV=production
- PORT=8088
- MAX_UPLOAD_SIZE_MB=500
- ENABLE_LOCAL_AI_OCR=true
- ENABLE_WHISPER_TRANSCRIPTION=true
- DATA_DIR=/app/data
volumes:
- ./data:/app/data
- ./uploads:/app/uploads
- ./exports:/app/exports
healthcheck:
test: ["CMD", "curl", "-f", "http://localhost:8088/health"]
interval: 30s
timeout: 10s
retries: 3Step 3: Secure Local Access with Reverse Proxy
To access SnapOtter over your local area network (or private WireGuard mesh) with HTTPS, bind your local reverse proxy (Caddy or Nginx) to the internal port 8088.
# Caddy Example for private mesh DNS
tools.internal.yourdomain.com {
tls internal
reverse_proxy 127.0.0.1:8088
}Step 4: Launch and Verify
Start the container in detached mode:
docker compose up -dVerify that the container is healthy and listening:
docker compose ps
curl -s http://127.0.0.1:8088/healthStep 5: Integrate Into Office Workflows
Once running, your team can access the web dashboard to drag-and-drop batches of PDF scans or audio recordings for immediate local processing. You can also script automated network directory watchers that ingest scanned receipts and automatically OCR them into searchable PDFs without employee manual labor.
Conclusion
Self-hosting tools like SnapOtter proves that modern productivity software doesn't require handing your company's data over to third-party cloud vendors. Local hardware running open-source software gives you total privacy, speed, and reliability.
Frequently asked questions
What is SnapOtter?
An on-premises file-processing tool: drag-and-drop batches of PDF scans or audio recordings for local AI OCR and transcription, with searchable-PDF exports and directory watchers that ingest scans automatically.
Why self-host document processing instead of using a cloud OCR service?
Zero outbound cloud calls. Client NDAs, HIPAA obligations, and proprietary documents stay on hardware you control, and there is no per-page metering on a vendor's API.
What does the install actually need?
Docker Compose binding port 8088 to localhost, volumes for data, uploads, and exports, and a local reverse proxy (Caddy or Nginx) for HTTPS access over the LAN or a private WireGuard mesh. A healthcheck endpoint confirms it is live.
