Veradoc Help Center

Introduction

Veradoc is a private Retrieval-Augmented Generation (RAG) platform designed for small and medium teams. It lets your team query their own documents using AI — without any data ever leaving your infrastructure.

Unlike cloud-based AI tools, Veradoc runs entirely inside Docker containers on your own machine or server. There is no telemetry, no cloud sync, and no third-party access to your documents.

Requirements

Veradoc has minimal requirements. The only thing you need installed is Docker Desktop.

Installation

Installation is a single command. Open a terminal (or PowerShell on Windows) and run:

Linux & macOS

# Run this single command in your terminal
$ curl -fsSL https://veradoc.ai/install.sh | bash

Windows (PowerShell as Administrator)

# Run in PowerShell (Admin)
PS> irm https://veradoc.ai/install.ps1 | iex

Once installation completes, open http://localhost:4200 in your browser. Veradoc will be running and ready to use.

1. 1
   Install Docker DesktopDownload from docker.com and follow the installer. No configuration needed.
2. 2
   Run the install commandOpen your terminal and paste the command above. The script pulls all necessary images and starts the services automatically.
3. 3
   Open your browserNavigate to http://localhost:4200. Create your admin account on the first launch screen.

Your First Authentication

After install VeraDoc. The system come with an Administrator role default account to start login. These are the credentials

1. Email: admin@veradoc.ai
2. Password: password

Your First Query

After logging in for the first time, follow these steps to upload a document and ask your first question:

1. Click New Collection in the sidebar and give it a name (e.g. "Contracts").
2. Drag and drop one or more documents into the collection. Supported formats include PDF, DOCX, TXT, and Markdown.
3. Wait for the green Indexed status — this means Veradoc has processed and embedded your documents.
4. Open the Chat tab, select your collection, and type your question.

LLM Models

Veradoc is model-agnostic. You can connect any OpenAI-compatible endpoint, a local Ollama instance, or a HuggingFace model.

Supported backends

Configure your model under Settings → AI Model after logging in. You can switch models at any time without rebuilding your stack.

Document Upload

Veradoc processes documents by splitting them into chunks, generating vector embeddings, and storing them locally in its built-in database. Supported file types:

• PDF — including scanned PDFs with OCR support
• DOCX / DOC — Microsoft Word documents
• TXT / Markdown — plain text and structured notes
• CSV — tabular data (experimental)

Team & Roles

Veradoc supports multi-user setups with role-based access control. There are three built-in roles:

Invite team members under Settings → Team. You can use Google OAuth or email/password depending on your configuration.

FAQ

Does Veradoc send any data to the cloud?

No. Veradoc runs entirely inside Docker containers on your own machine or server. No documents, queries, or responses are sent anywhere outside your infrastructure.

Can I use Veradoc without an internet connection?

Yes, after the initial installation. The first run requires internet access to pull Docker images. After that, Veradoc operates fully offline — especially if you use a local model like Ollama.

How do I stop or restart Veradoc?

# Stop all services
$ curl -fsSL https://veradoc.ai/scripts/stop.sh | bash

# Start again
$ curl -fsSL https://veradoc.ai/scripts/start.sh | bash

Is my data lost when I stop Veradoc?

No. All documents and embeddings are stored in a persistent Docker volume. Stopping or restarting the containers does not delete your data.

What happens when I update Veradoc?

Run the install command again — it will pull the latest images and restart services while preserving all your data.

Troubleshooting

Port 4200 is already in use

Another application is using port 4200. You can change the port by editing the docker-compose.yml file in the veradoc/ directory and replacing 4200:4200 with YOUR_PORT:4200.

The embedding model takes a long time to load

On first run, the embedding model is downloaded (typically 200–500 MB). This is a one-time operation. Subsequent starts are much faster.

I can't connect to the API server

Check that Docker is running and all containers are healthy:

$ docker compose ps

All services should show Up status. If any show Exit, check logs with docker compose logs <service-name>.

Still stuck?

Open an issue on GitHub or email veradocai@gmail.com. Include the output of docker compose logs to help us diagnose the issue faster.