Inspiration
student monitoring and early intervention is very challenge
What it does
EduGuard is an intelligent student monitoring and early intervention routing system. It integrates fragmented data across student information systems (SIS), learning management systems (LMS), and Special Education platforms.
How we built it
Components Agent Registry & Gateway (Google ADK Sim): Catalogues active agents (specifications, roles, schemas) and handles routing. Enforces Model Armor guardrails on all outgoing LLM completions. Attendance Anomaly Agent: Evaluates SIS attendance records (absences, tardies) and LMS grade logs (grades < 70%, late submissions) to compute risk metrics and flag at-risk students. IEP Compliance Agent: Retrieves Special Education records to verify if a student's IEP accommodations (e.g. extended time, calculators) are properly delivered in classroom assessments. Intervention Routing Agent: Synthesizes academic findings and IEP reviews to draft a comprehensive counselor support plan template using LLMs (Gemini / GPT) or a fallback high-quality template generator. Model Armor (Vertex AI simulation): Interceptors at the Gateway redact student names, guardian names, contact numbers, email addresses, and DOBs before they exit the secure database zone. Triggers an authorized re-hydration flow once the counselor reviews and approves the intervention plan. OpenTelemetry Logging: Emits structured trace spans for agent-to-agent communication, visualizing parent/child relationships and execution latency directly in the dashboard UI. Cloud SQL Simulation: A SQLite local database populated with student demographics, recent class grades, attendance histories, and special education requirements.
Challenges we ran into
kind of challenge to expand it
Accomplishments that we're proud of
baseline project
What we learned
Antigravity is very powerful
What's next for EduGuard AI & Privacy
continue enhancement
Log in or sign up for Devpost to join the conversation.