curiosityai/curiosity-non-root
Build enterprise AI on your data, in your environment.
50K+
curiosityai/curiosity-non-root repository overview
Curiosity Studio
Build enterprise AI on your data, in your environment.
Curiosity Studio is a platform for building internal AI applications, dashboards, search experiences, and agents. Connect your data, describe the tool you need, and give people and AI models the context to work with it.
The curiosityai/curiosity image packages Studio’s knowledge graph, search, and AI runtime in one container. Run it locally to develop and evaluate, or deploy it on your own infrastructure.
What you can do with Studio
- Build internal tools from a description, grounded in your connected data.
- Search across systems using keyword search, vector search, and graph traversal.
- Give AI agents context from a typed knowledge graph of entities and relationships.
- Connect existing data sources, including SharePoint, Teams, Slack, Confluence, Jira, Salesforce, and S3.
- Respect source permissions when retrieving data and generating answers.
- Use your choice of model, from commercial APIs to models hosted on your own infrastructure.
- Build with code using REST APIs and C# SDKs.
Use Studio for applications such as research agents, support assistants, review dashboards, and compliance trackers.
Quick start
Get a local Studio instance running with Docker.
Prerequisites
- Docker installed and running
- At least 8 GB of RAM available to Docker; 16 GB is recommended when working with real data
- Port
8080available - OpenSSL for the password-generation command below
Start the container
On Linux or macOS:
mkdir -p ~/curiosity/storage
ADMIN_PASSWORD=$(openssl rand -base64 24)
echo "Initial admin password: $ADMIN_PASSWORD"
docker run --name curiosity \
-p 127.0.0.1:8080:8080 \
-v ~/curiosity/storage:/data \
-e MSK_GRAPH_STORAGE=/data/curiosity \
-e MSK_ADMIN_USER=admin \
-e [email protected] \
-e MSK_ADMIN_PASSWORD="$ADMIN_PASSWORD" \
curiosityai/curiosity:latest
Open http://localhost:8080 and sign in with username admin and the generated password. First startup may take a few moments while the database and indexes initialize.
The volume keeps Studio’s graph storage outside the container. The port binding makes this example accessible only from your own machine.
Local development only: This example does not configure TLS. For a shared or production deployment, follow the Docker deployment guide and the production guidance linked from the quickstart.
Admin account configuration
Set MSK_ADMIN_USER, MSK_ADMIN_EMAIL, and MSK_ADMIN_PASSWORD together. These values are applied each time the container starts, so a password changed in the UI will be reset on restart while the variables remain configured.
For Windows commands and the next steps after signing in, see the full quickstart.
Image tags
| Tag | Description |
|---|---|
latest | Most recent build; updated on every build. |
nightly | Tracks the same most recent build as latest. |
| Numeric build ID | A specific, immutable build, such as 66474. |
| Calendar-versioned release | An official release, such as 26.6.66474. |
For staging and production, pin a numeric build or release tag so upgrades are deliberate. See the Docker deployment guide for details.
Deployment and configuration
Studio can run on premises or in a private cloud. For shared deployments, plan for persistent storage, authentication, backups, and a TLS-terminating reverse proxy.
Pricing and resources
The developer version is free for development and evaluation. Enterprise Studio is available for organizations that need a tailored deployment and support. See pricing and plans.
Tag summary
Content type
Image
Digest
sha256:4e80780a4…
Size
2.2 GB
Last updated
about 2 hours ago
docker pull curiosityai/curiosity-non-root