Ertas for Education

Fine-tune and deploy AI models on curriculum and student interaction data within your institution's own infrastructure, keeping student records private and FERPA-compliant.

The Challenge

Educational institutions at every level — from K-12 districts to research universities — are under pressure to personalize learning at scale. Adaptive tutoring, automated grading, curriculum recommendation, and content generation all benefit from AI, but the models need to understand the specific curriculum, pedagogical approach, and assessment standards of the institution. Generic language models produce plausible-sounding but often incorrect explanations, misalign with grade-level expectations, and lack awareness of specific course syllabi or institutional learning outcomes.

Student data privacy adds a formidable constraint. FERPA imposes strict requirements on how educational records are handled, and parents, students, and faculty are increasingly skeptical of sending student interaction data to third-party cloud AI providers. Many school districts have outright banned commercial AI tools that transmit student data externally. Institutions that want to use AI for grading assistance or tutoring must find a way to keep student data within their own systems while still accessing the benefits of modern language models.

The Solution

Ertas gives educational institutions a complete pipeline for building curriculum-aware AI models that run entirely within their own infrastructure. With Ertas Studio, instructional designers and ed-tech teams can fine-tune foundation models on institution-specific content — course materials, past exam questions, rubrics, and anonymized student interaction logs — using LoRA adapters for efficient training. The resulting models understand the institution's terminology, assessment criteria, and pedagogical style in ways that off-the-shelf models cannot.

Deployment happens on-premise or within the institution's private cloud. Ertas Cloud provisions private inference endpoints that integrate with existing LMS platforms, and Ertas Vault ensures that all training data and student interaction logs are encrypted, access-controlled, and retained only as long as institutional policy requires. Faculty and administrators get AI-powered tools — adaptive practice question generators, automated first-pass grading, and personalized study recommendations — without ever exposing student records to external services.

Key Features

Studio

Curriculum-Aware Fine-Tuning

Use Studio's visual canvas to fine-tune models on JSONL datasets of course content, assessment rubrics, Q&A pairs, and pedagogical guidelines. LoRA adapters let you create subject-specific models — one for introductory biology, another for AP calculus — without full retraining costs.

Hub

Educational Model Library

Browse Hub for community-contributed educational base models and adapters — including models pre-trained on open textbook corpora, science reasoning datasets, and multilingual educational content — to accelerate your institution's fine-tuning efforts.

Cloud

LMS-Integrated Endpoints

Deploy fine-tuned models as private Cloud endpoints that integrate directly with Canvas, Moodle, Blackboard, or custom LMS platforms. Endpoints run within your institution's network, keeping all student interaction data local and audit-ready.

Vault

FERPA-Aligned Data Protection

Vault encrypts all training data and inference logs at rest and in transit, enforces role-based access controls for faculty, staff, and administrators, and provides configurable retention policies that align with FERPA and institutional data governance requirements.

Example Workflow

A large university's online learning team wants to provide adaptive tutoring for its introductory computer science course, which enrolls 2,000 students per semester. The team exports 15,000 anonymized student Q&A interactions from the course forum, along with the complete course textbook and assignment rubrics, as a JSONL dataset and uploads it to Ertas Vault. In Ertas Studio, they select a Phi-3 base model from Hub and fine-tune it with a LoRA adapter targeting Socratic tutoring — training the model to guide students toward answers through questions rather than providing direct solutions. After two hours of training, the model is deployed as a private endpoint within the university's data center via Ertas Cloud and connected to the course's Moodle instance. Students interact with the AI tutor through the familiar LMS interface, receiving personalized hints and explanations that align with the course's specific curriculum and grading rubric. All interaction data stays within the university's infrastructure, and the teaching team can review aggregated usage analytics through Vault's audit dashboard without exposing individual student records.

Compliance & Security

Ertas supports FERPA-aligned deployments by keeping all student data and inference processing within institution-controlled infrastructure. Vault's encryption, access controls, and retention policies help institutions meet FERPA's requirements for protecting educational records. Institutions are responsible for their own FERPA compliance assessments and for ensuring proper anonymization of student data before use in model training.