Dietrich Analysis Research Education Platform
A Complete Ecosystem for AI Integration in Academic Settings
Dietrich Analysis Research Education (DARE) is a comprehensive platform that functions as a "large language model (LLM) server starter pack" for higher education. DARE provides direct application programming interface (API) access to LLMs while maintaining a focused, accessible interface that supports progressive skill development and emphasizes transparency and user agency in human-AI collaboration.
DARE Platform Architecture & Capabilities
Dashboard
Comprehensive usage analytics and activity tracking
Multi-Model Access
Direct API access to Anthropic Claude models, OpenAI models and emerging open-source alternatives
Document Management
Secure file storage with advanced Retrieval Augmented Generation (RAG) capabilities
System Prompts
Customizable prompt templates for consistent AI interactions
Workflow Automation
Multi-step automated processes combining prompts, documents and models
Cost Tracking
Detailed token usage monitoring and billing transparency
Take a Look
Click images to enlarge
Learning Progression Framework
DARE supports natural skill development from basic system prompts to sophisticated multi-agent implementations, enabling users to progress organically through:
- System prompt construction and parameter adjustment
- Data integration through RAG workflows
- Multi-agent tool exploration and customization
- Innovation and custom application development
Research-Driven Development
Humanities and social sciences faculty across CMU's Dietrich College actively conduct research and experiments to validate effective AI integration approaches while maintaining academic integrity. This research-driven innovation pipeline combines tool development with empirical validation, ensuring DARE's evolution remains grounded in demonstrated educational value and measurable outcomes.
Strategic Partnership and Open Source Mission
DARE operates as the technical prototype partner for the Open Forum for AI (OFAI), a university-led collaborative initiative shaping human-centered, responsible and transparent AI implementation. This partnership reflects a commitment to democratizing advanced AI capabilities in academic settings through open-source development and low deployment costs, making sophisticated tools accessible to institutions regardless of resources.
DARE was developed by Dietrich Computing and Operations, led by Vincent Sha, associate dean for IT and operations.