This guide walks you through setting up Dittah Studio for the first time. By the end, you'll have a running instance connected to your data sources, ready to build your first workflow or start chatting with your documents.
Prerequisites
Before you begin, make sure you have:
- Docker installed and running (Docker Desktop on Mac/Windows, or Docker Engine on Linux)
- At least 8 GB of RAM allocated to Docker
- A modern web browser (Chrome, Firefox, Safari, or Edge)
Step 1: Download & Start Dittah Studio
Head over to the Dittah Studio GitHub repository for prerequisites, download instructions, and startup commands. The repo README walks you through everything you need to get the containers running.
Once the containers are up, open your browser and navigate to http://localhost.
Step 2: Initial Configuration
On first launch, you'll see the setup wizard. Here's what to configure:
- Admin account — Create your administrator username and password.
- LLM provider — Dittah uses a large language model only during workflow design — never at runtime. You have two approaches depending on your data sovereignty needs:
- Fully on-premise (private GPT) — Use Ollama for both LLM and embeddings. Your data never leaves your network. This is the recommended setup for air-gapped or highly regulated environments.
- Hybrid / cloud — Use a cloud LLM API such as OpenAI, Google Gemini, Groq, or Anthropic. You can also use CLI tools like Claude Code or Gemini CLI that route through your own API keys. This gives you access to the most capable models while Dittah still runs entirely on your infrastructure.
In short: Ollama = completely local, no external calls. Cloud APIs = more powerful models, but LLM requests leave your network during workflow design only. Pick the approach that matches your organization's requirements.
Step 3: Explore the Platform
Dittah Studio has three main areas:
- Build Studio — Create data workflows using natural language. Describe what you need, test the generated code, then freeze it for production.
- Chat Studio — Upload documents and ask questions. Get instant answers with source citations and role-based access control.
- Ops Hub — Monitor running workflows, view execution logs, and troubleshoot issues.
Step 4: Connect Your Data
Dittah supports a wide range of data sources out of the box:
- Files — Excel, CSV, JSON, XML, PDF, Word, PowerPoint
- Transfer protocols — SFTP, FTP, email (IMAP)
- APIs — REST endpoints with authentication
- Databases — PostgreSQL, MySQL, SQL Server
What's Next?
Now that your environment is set up, try these next steps:
- Build your first workflow — a step-by-step tutorial
- Understanding Data Fences — configure role-based access (coming soon)
- Compare plans — see what's included in Community vs. Professional
Need help? Contact our team or open an issue on GitHub.
