For Education¶
RosettaLabs -- virtual labs at scale with institutional guardrails.
Overview¶
RosettaLabs enables universities and training programmes to deliver cloud-based lab environments to hundreds of students without manual provisioning. Students never need to see a cloud console, create cloud credentials, or understand infrastructure. Instructors create a formation once; students launch it with a single click. Budget controls, SSO integration, and spot instance support keep costs predictable and administration minimal.
The MetaCloud handles all the underlying complexity -- provisioning instances, configuring networking, managing credentials, mounting storage -- so that both instructors and students can focus entirely on teaching and learning.
RosettaHub has been serving universities since 2016.
Migrating from Azure Lab Services?
Microsoft is retiring Azure Lab Services on June 28, 2027. RosettaHub is a purpose-built alternative with more flexibility, lower costs, and multi-cloud support. See the migration guide.
Key Capabilities¶
Virtual Labs at Scale¶
Deliver consistent, pre-configured environments to every student in a course. Labs launch immediately -- there is no publishing step or image creation delay. Supported lab types include:
| Lab Type | Description |
|---|---|
| Jupyter Notebook | Python, R, Julia data science environments |
| RStudio | Statistical computing and visualization |
| VS Code | Browser-based development environment |
| Linux Desktop | Full graphical desktop via browser (Virtual Lab formation) |
| Windows Desktop | Windows applications accessible through the browser |
| Docker Formation | Custom containerized applications and toolchains |
| Big Data Cluster | Spark and Hadoop on AWS EMR or GCP Dataproc |
| HPC Cluster | High-performance computing with AWS ParallelCluster |
| Kubernetes Cluster | Container orchestration for advanced courses |
Each lab type is backed by a Formation -- a cloud-agnostic template that deploys identically on AWS, Azure, GCP, or any other connected provider.
Full Cloud Flexibility¶
Unlike managed lab services that restrict users to a handful of fixed configurations, RosettaHub gives educators the full flexibility of the underlying cloud providers:
| Capability | What This Means |
|---|---|
| Any instance type | Choose from the complete catalog of any cloud provider -- not a limited set of 4-8 options |
| Any image size | No artificial cap on disk size |
| Any cloud provider | AWS, Azure, GCP, OVH, OCI, and OpenStack from one platform |
| Stop and restart | Students can stop VMs and restart them later without losing work -- no forced termination |
| Per-second billing | Pay for the exact time a VM runs, down to the second -- not rounded up to the nearest hour |
| IDE and notebook support | All IDEs and notebooks are supported, including Jupyter, RStudio, Zeppelin, and VS Code |
Container-Based Labs¶
Start labs from pre-built container images with popular tools ready to use:
- Jupyter notebooks with Python, R, and Julia kernels
- RStudio with R packages pre-installed
- VS Code in the browser via code-server
- Custom Docker images -- build your own and share across courses
Container-based labs use Docker Formations and launch in seconds. Instructors can customize container images, add packages, and share the result with students -- no image publishing required.
One-Click Student Experience¶
Students receive a self-service portal scoped to their course. They do not need cloud accounts, CLI tools, or infrastructure knowledge.
- Student logs in via institutional SSO
- Pre-approved formations appear in their dashboard
- Student clicks Launch -- the environment is ready in minutes
- Work is persisted through snapshots and shared storage
Dedicated Storage per Student¶
Each student can have a dedicated cloud storage (object storage or file system) automatically mounted to their lab environment. This ensures student work is saved independently of the VM lifecycle -- even if a spot instance is reclaimed or a VM is deleted, the student's files are preserved on their personal storage.
Formation Templates for Instructors¶
Instructors follow a straightforward workflow:
- Create or clone a formation with the required software stack
- Test the environment
- Share with the course organization -- all enrolled students gain access
- Optionally use Launch on Sharees to deploy environments for the entire class in one action
Labs can be launched immediately after creation -- no publishing step or image baking required. Instructors can also share formation templates with colleagues across departments or institutions.
Batch Operations
Formation batch actions -- Launch on Sharees, Stop on Sharees, Delete on Sharees -- allow instructors to manage an entire class of environments from a single menu.
Real-Time Collaboration¶
RosettaHub makes collaboration a first-class capability:
- Share formations with individual users, groups, or entire organizations
- Share lab templates between instructors across departments
- Share images -- students and instructors can share customized environments
- Cross-institutional sharing -- collaborate across organizational boundaries via URLs or marketplaces
Budget Controls¶
Spending is governed at every level of the organization hierarchy:
| Scope | Example |
|---|---|
| Department | Computer Science -- $10,000 per semester |
| Course | CS 301 Machine Learning -- $2,000 |
| Student | Individual allocation -- $50 |
Budgets are enforced in real time. When a student's allocation is exhausted, new launches are blocked -- no surprise bills at semester end.
Cost Optimization¶
Per-Second Billing¶
RosettaHub bills at actual cloud provider prices with per-second granularity. Unlike managed lab services that charge per hour at multiples of infrastructure cost, students pay only for what they use. A 20-minute lab session costs 20 minutes, not a full hour.
Spot Instances¶
Spot and preemptible instances provide 60-90% cost savings on compute. RosettaHub handles spot interruptions automatically:
- Snapshot on Termination -- student work is preserved if a spot instance is reclaimed
- Hibernate -- suspend instances and resume later without re-provisioning
- Auto-replace -- automatically launch a replacement VM with similar characteristics
- Fallback to On-Demand -- critical lab sessions can automatically switch to standard pricing
Automated Shutdowns¶
- Scheduled shutdowns -- automatically stop all labs at the end of a class period
- Idle VM shutdown -- stop VMs that have been idle for a configurable period
Dedicated Cloud Accounts¶
Each user operates within a dedicated cloud account (AWS sub-account, Azure resource group, or GCP project) with independent quotas. This eliminates the shared-quota bottleneck that plagues managed lab services -- instructors never hit scaling limits when launching VMs for a large class, because each account has its own resource quotas from the cloud provider.
Direct Cloud Access for Cloud Learners¶
The MetaCloud simplifies cloud access for students who don't need to learn infrastructure. But for cloud computing courses where students need to learn AWS, GCP, or Azure directly, RosettaOps provides federated cloud console access -- students get direct access to native cloud consoles within their sandboxed accounts.
This is essential for:
- Cloud certification courses -- students work in the real AWS/GCP/Azure console, not a simulation
- AI and ML courses -- students access native services like SageMaker, Bedrock, Vertex AI, and Azure OpenAI directly
- DevOps and platform engineering -- students learn real IAM, networking, and infrastructure tools
- Data engineering -- students use native services like Athena, BigQuery, and Redshift
RosettaOps ensures students cannot exceed their budget or escape their sandbox, even with full native console access. Educators get the governance guardrails; students get the real cloud experience.
Single Management Portal¶
Instructors and administrators manage everything from one portal -- no switching between separate consoles for lab plans, lab creation, user management, and billing. Organization hierarchy, budgets, formations, machines, and users are all accessible from the same interface.
SSO Integration¶
Connect your institutional identity provider for seamless access:
- SAML 2.0 -- Shibboleth, ADFS, Azure AD
- LDAP -- Active Directory, OpenLDAP
- OAuth -- Google Workspace, GitHub, custom providers
Next Steps¶
- Migrate from Azure Lab Services -- migration guide with feature comparison
- Quick Start -- set up your RosettaHub account
- Formations -- learn how to create lab templates
- Organizations -- structure courses and departments
- Tutorials -- step-by-step walkthroughs