Built for Institutions
Security & Compliance
- Designed for FERPA compliance — full audit trail of who accessed what and when. Institutions should conduct their own FERPA assessment for their specific deployment.
- Student identifiers stripped from all AI prompts. With Ollama (on-premise), submission data stays on your network. With cloud AI providers (Anthropic, OpenAI), submission content is transmitted for processing.
- Three AI deployment options: Anthropic Claude, OpenAI, or Ollama (open-source local AI runtime — on-premise, no cloud dependency)
- Per-instructor data isolation — faculty cannot view each other's student data
- Consent-based admin access with time-limited approvals and revocation
- Automatic temp file cleanup after every grading job
- Job deletion on demand for complete data purge
Enterprise Ready
- Multi-tenant architecture with custom branding per department or institution (logo, colors, CSS)
- SSO via Azure Entra ID (Active Directory). Configuration interface for SAML 2.0, Google Workspace, and generic OIDC — implementation in progress.
- Role hierarchy: Platform Admin → Org Admin → Instructor → Teaching Assistant
- 430+ REST API endpoints with interactive Swagger documentation for integrations
- Per-module AI model selection — match model capability to task complexity
- Configurable usage quotas and real-time usage analytics per user and department
- Bring your own API keys or use org-provided keys — full control over AI access
What's Built
Grading & Assessment
- ✓ 9 auto-graders: SQL, Essay, Tableau, Excel, Python, R, Power BI, Access, Statistics
- ✓ D2L Brightspace ZIP import on all graders
- ✓ PDF feedback reports with institutional branding
- ✓ D2L-compatible gradebook CSV export
- ✓ AI Oral Exams with adaptive follow-up questions
- ✓ Peer Evaluation with anonymous scoring
- ✓ Plagiarism detection and AI vulnerability scoring
Content & Platform
- ✓ 5 assignment generators with answer keys and instructor guides
- ✓ Virtual Teaching Assistant, Accessibility Checker
- ✓ Presentation Generator, Lecture Studio, CellStream
- ✓ Multi-tenant SaaS with per-department org branding
- ✓ 430+ REST API endpoints with interactive documentation
- ✓ SSO via Azure Entra ID; SAML 2.0, Google Workspace, OIDC configuration in progress
- ✓ 3 AI provider options: Anthropic, OpenAI, Ollama (local)
Testing & Validation
476 automated tests covering API routes, grader logic, auth, and data isolation
4 AI models benchmarked across 8 grading task types with deterministic scoring
FERPA audit complete — 6 HIGH findings identified and remediated across 430+ routes
Ready for faculty beta testing. Contact [email protected] to schedule a live demo with your assignment type.
Agoge Development LLC
Founded by educators,
for educators.
About Agoge Development
Agoge Development was founded by educators who experienced firsthand the gap between what technology promises and what actually works in a real classroom. We build tools that solve the problems instructors actually have — starting with the hours lost to grading and content creation.
The name Agoge comes from the ancient Spartan education system, renowned for its emphasis on excellence through rigorous, purposeful training. We believe every instructor deserves tools worthy of that standard.
The Agoge Platform is built for data analytics programs in higher education — purpose-designed for the tools your students actually use: SQL, Python, R, Excel, Tableau, and Power BI. It's not a general-purpose LMS plugin. It's a specialist toolkit built around your workflow.
Early access available now — individual instructors, departments, and institutions welcome. Request a demo at agogedev.com or write to [email protected].
AI-assisted grading is designed as an instructor aid. Results should be reviewed by the instructor before release to students. The Agoge Platform is a teaching tool, not a replacement for professional academic judgment.