Academic Research Computing Platform - Pre-configured cloud environments made simple
Prism provides researchers with pre-configured cloud workstations for data analysis, machine learning, and computational research. Launch production-ready environments without manual configuration.
From individual researchers to institutional deployments - research computing made simple, scalable, and cost-effective.
Learn more at prismcloud.io
- 🎯 Default to Success: Every template works out of the box in every supported region
- ⚡ Optimize by Default: Smart instance sizing and cost-performance optimization
- 🔍 Transparent Fallbacks: Clear communication when configurations change
- 💡 Helpful Warnings: Gentle guidance for optimal choices
- 🚫 Zero Surprises: Users always know what they're getting
- 📈 Progressive Disclosure: Simple by default, detailed when needed
Homebrew (Recommended)
brew install scttfrdmn/tap/prismManual Installation
# Download and extract
curl -L https://github.com/scttfrdmn/prism/releases/latest/download/prism_0.5.11_darwin_arm64.tar.gz | tar xz
# Install binaries
sudo mv prism prismd /usr/local/bin/Debian/Ubuntu
wget https://github.com/scttfrdmn/prism/releases/download/v0.5.11/prism_0.5.11_linux_amd64.deb
sudo dpkg -i prism_0.5.11_linux_amd64.debRHEL/CentOS/Fedora
wget https://github.com/scttfrdmn/prism/releases/download/v0.5.11/prism_0.5.11_linux_amd64.rpm
sudo rpm -i prism_0.5.11_linux_amd64.rpmAlpine Linux
wget https://github.com/scttfrdmn/prism/releases/download/v0.5.11/prism_0.5.11_linux_amd64.apk
sudo apk add --allow-untrusted prism_0.5.11_linux_amd64.apkScoop
scoop bucket add scttfrdmn https://github.com/scttfrdmn/scoop-bucket
scoop install prismManual Installation
# Download from GitHub releases
# https://github.com/scttfrdmn/prism/releases/latest
# Extract and add to PATHLaunch your first workspace with a guided wizard!
# Configure AWS credentials (if not already set up)
aws configure
# Launch the interactive wizard
prism initThe wizard will guide you through:
- ✅ Template Selection: Browse by category (ML/AI, Data Science, Bioinformatics, Web)
- ✅ Workspace Configuration: Name and size with cost estimates
- ✅ Review: See exactly what you're launching
- ✅ Launch: Real-time progress tracking
- ✅ Connect: SSH connection details and next steps
Time to first workspace: ~30 seconds 🚀
For experienced users or automation:
# View available templates
prism templates
# Launch a Python ML environment
prism launch python-ml my-research
# Connect via SSH
prism connect my-research
# View running workspaces
prism listAutomatic Features:
- ✅ Daemon starts if not running
- ✅ Optimal instance type selected
- ✅ Security groups configured
- ✅ SSH keys generated and managed
- ✅ Template provisioned and ready
Credential Discovery (automatic):
- Environment variables (AWS_PROFILE, AWS_ACCESS_KEY_ID)
- AWS CLI configuration (~/.aws/credentials)
- Prism profiles (for multi-account management)
- Hibernation: Preserve state while reducing costs by 90%
- Idle Detection: Automated hibernation policies with configurable thresholds
- Budget Management: Project-level cost tracking and alerts
- Cost Analytics: Real-time spending reports and forecasts
- 21+ Pre-configured Environments: Python ML, R, bioinformatics, web dev, and more
- Template Inheritance: Compose complex environments from simple building blocks
- Smart Defaults: Optimal instance sizing and cost-performance ratios
- Regional Fallbacks: Automatic handling of availability constraints
- Project-Based Organization: Multi-user projects with role-based access
- Research User System: Persistent identities across workspaces
- Multi-Account Support: Manage multiple AWS profiles efficiently
- Template Marketplace: Share and discover community templates
- CLI: Fast, scriptable command-line interface
- TUI: Interactive terminal interface with keyboard navigation
- GUI: Desktop application (available when building from source)
- REST API: Complete HTTP API on port 8947
Prism includes 21+ pre-configured templates for research computing:
- Python ML: Jupyter, scikit-learn, TensorFlow, PyTorch
- R Research: RStudio, tidyverse, Bioconductor
- Bioinformatics: BLAST, bowtie2, samtools, bedtools
- Web Development: Node.js, Docker, nginx
- Deep Learning: GPU-optimized environments with CUDA
# View all templates
prism templates
# Get detailed template info
prism templates info python-ml# Launch a workspace
prism launch python-ml my-project
# List running workspaces
prism list
# Connect via SSH
prism connect my-project
# Stop workspace
prism stop my-project# Hibernate to preserve state while saving costs
prism hibernate my-workspace
prism resume my-workspace
# Automated idle policies
prism idle profile list
prism idle workspace my-gpu --profile gpu# Create project with budget
prism project create ml-research --budget 500
# Add team members
prism project member add ml-research [email protected] --role member
# Launch workspace in project
prism launch python-ml analysis --project ml-research# Command line
prism templates
# Terminal UI
prism tui
# REST API
curl http://localhost:8947/api/v1/instances📚 Complete Documentation Site - User guides, architecture docs, and persona walkthroughs
prism --help # Show all commands
prism templates # List available templates
prism templates info <template> # Detailed template info
prism doctor # System health checkGuides:
- AWS Setup Guide - AWS account and credential configuration
- Installation Guide - Comprehensive installation instructions
- Budget System Philosophy - Multi-budget system design and conceptual model (v0.5.10+)
- Budget Banking Philosophy - Surplus tracking and burst budgeting
- Resource Tagging - Cost optimization and zombie resource cleanup
- Security & Compliance - NIST 800-171, HIPAA, GDPR, FISMA compliance
- Changelog - Version history and release notes
- User Invitation System: Individual, bulk, and shared token invitations with full lifecycle management
- Automatic Provisioning: Research users created with SSH keys, UID/GID, and EFS home directories on invitation acceptance
- Quota Validation: Pre-flight AWS capacity checking prevents bulk invitation failures
- Professional GUI: Cloudscape-based invitation management interface with QR code generation
- Zero Manual Setup: End-to-end automation from invitation send to workspace access
- Multi-Project Budgets: Projects can reference multiple budgets for complex funding scenarios
- Budget Surplus Banking: Track and leverage surplus funds for burst research needs
- Budget Performance Metrics: ROI analysis, utilization tracking, and forecasting
- Complete Prism Rebrand: CloudWorkStation → Prism across entire codebase (29,225 files)
- Binary Rename:
cws/cwsd→prism/prismd
- Streamlined Navigation: 14 → 6 top-level menu items for clearer workspace focus
- Unified Storage: Single storage interface combining EFS and EBS management
- Hierarchical Settings: Advanced features organized under collapsible settings
- Terminal/WebView Integration: Merged into workspaces view for better context
- Quick Start Wizard: Launch first workspace in 30 seconds with GUI wizard
- CLI Init Command: Interactive
prism initonboarding in terminal - Workspace Terminology: Consistent "workspace" naming across all interfaces
- Background State Monitoring: Async daemon monitoring of AWS state changes
- Billing Accuracy: Correct hibernation billing exception handling
- Reliability Improvements: AWS system status checks, IAM eventual consistency
- Template File Provisioning: Provision files directly from template definitions
- AWS Cost Testing: Complete integration test suite with cost tracking
- Instance Lifecycle Testing: Full workflow validation including cleanup
- Project Rename: CloudWorkStation → Prism (complete rebrand)
- Repository Rename:
cloudworkstation→prismon GitHub - Configuration Directory:
.cloudworkstation→.prism - Module Path Update: Complete Go module path migration
- EMR Studio: Big data analytics and Spark-based research
- Amazon Braket: Quantum computing research access
- Web Service Framework: Unified interface for EC2 + AWS research services
- Dynamic OS Versions: Choose OS versions at launch time with
--versionflag - Version Aliases: Support for
latest,lts,previous-lts - AMI Freshness Checking:
prism ami check-freshnessvalidates static AMI IDs - AWS SSM Integration: Automatic latest AMI discovery for major distributions
- Package Management: Available via Homebrew (macOS), Scoop (Windows), deb, rpm, apk
- Multi-User Architecture: Persistent research identities across workspaces
- SSH Key Management: Complete key generation and distribution
- Template Registry: Multi-registry support with community templates
- Policy Framework: Institutional governance and access control
- Project-Based Organization: Multi-user projects with role-based access
- Budget Management: Real-time cost tracking and automated controls
- Hibernation Ecosystem: Manual + automated idle detection policies
- Template Inheritance: Stackable template system
Phase 5 (Current): Multi-user collaboration and template marketplace Phase 6: Advanced storage (FSx, S3 integration) and AWS research services Phase 7: Enterprise authentication (OAuth, LDAP, SAML) and TUI enhancements
Prism is open source and welcomes contributions!
- Issues: Report bugs or request features
- Pull Requests: Submit code improvements
- Templates: Contribute research environment templates
- Documentation: Help improve guides
Development:
git clone https://github.com/scttfrdmn/prism.git
cd prism
make build
make testApache License 2.0 - Free for academic and commercial use
- Documentation: Complete docs site or
prism --help - System Check:
prism doctor - Issues: GitHub Issues
- Discussions: GitHub Discussions
- AWS Setup: See AWS Setup Guide
Prism v0.5.11 - Research computing environments made accessible | prismcloud.io