George Cann

DARE

Co-created Carnegie Mellon's open-source AI platform for education. 2,000+ students served in courses from history to statistics, with partner institutions in the US and abroad piloting it.

• One governed gateway to OpenAI, Anthropic, Gemini, Llama, Mistral and other open-source models, with role-based access and SSO.
• RAG grounded in course materials, a visual workflow builder for multi-step teaching tools, and a research mode.
• Usage, cost and per-message energy tracking, with audit trails for transparency.
• Flagship of CMU's Open Forum for AI: open source (AGPL 3.0) and self-hosted with Docker, so any institution can run it on its own terms.

Exit Tickets

After-class feedback platform: several hundred student submissions a week across multiple courses.

• Compares student feedback with each lecture's learning goals; custom reports flag action items and gaps in understanding for instructors.
• Gives students real-time feedback on their responses as they write, which has greatly improved response quality.
• Homework: students submit work and are assigned peer grading, which is compared against AI grading.

fieldwork.cloud

AI platform for field data and document research. Private beta.