EdConnect: The Ethical, Research-Backed AI Assistant for Educators.

Why Use a Generic Chatbot When You Can Have a Curriculum Expert?

Transparency Statement

Radical Transparency: Specialized Model & Ethical Constraints

The Pain: Generic large language models (LLMs) cannot be trusted for curriculum design. They hallucinate standards, lack pedagogical context, and operate on undisclosed data. Teachers deserve an expert, not a generalist.

Our Philosophy: The solution is a specialized, constrained, and transparent model. EdConnect is the realization of the Pragmatic Idealist vision: a dedicated Instructional Technology Coach AI that operates solely within defined ethical and pedagogical boundaries.

The Tool: EdConnect is a custom-built small language model (SLM) grounded in over 100 peer-reviewed and primary source documents (ISTE, UDL, NGSS, etc.). It provides clear, actionable, and citable guidance on curriculum, pedagogy, and edtech, filling the gap for the overworked professional.

We are not a black-box corporation. We are an open-source research initiative. You deserve to know exactly how this tool works.

Base Model (SLM): Claude Sonnet 4.5 API (Custom-Built). This is a small, specialized language model (SLM) built through Chatbase, explicitly tuned to the role of a Curriculum and Instructional Technology Coordinator.

Curated Database: 100+ Peer-Reviewed and Primary Source Documents. The model is grounded in authoritative frameworks (UDL, ISTE, NGSS, CASEL, etc.) to ensure all advice is research-based and citable.

Research & Citing Engine: Tavily Web Search Integration. The model is instructed to search and cite reputable sources (ResearchGate, ERIC, JSTOR, etc.) to prevent fabrication and ensure Academic Integrity.

Privacy: No student data stored. Inputs are ephemeral. The model is constrained to educational topics only. Conversation are stored briefly through AWS servers and deleted monthly.


Common Questions & Ethical Commitments (FAQ)

  • EdConnect is not a generalist. It is a specialized, constrained model with a curated database of educational standards and frameworks. Its logic is hard-coded to prioritize pedagogy and cite sources, which generic LLMs cannot reliably do.

  • Yes, this tool is completely free. It is sustained by the open-source community and the founder's commitment to ethical AI in education, currently operating at an explicit financial loss to ensure its accessibility.

  • The model is instructed to use a Research & Citing Engine to verify claims against reputable sources like ERIC and ResearchGate. If a source cannot be verified, it is flagged as 'anecdotal observation.'

  • We collect no personally identifiable information (PII) as there is no user login. Conversations are stored on secure AWS servers for operational purposes (e.g., monitoring popular topics and improving the model's educational focus), but all conversation logs are permanently deleted every 30 days. We never store, track, or build profiles.