# University of Wyoming IEP Platform: AI-Powered Educational Planning

Building a scalable, AI-driven platform for creating Individualized Education Programs with specialized agents and real-time collaboration.

70%

IEP creation time reduction

100+

Concurrent users supported

6

Specialized AI agents deployed

## The Challenge

Creating Individualized Education Programs (IEPs) is one of the most time-consuming tasks in special education. Teachers and administrators spend hours drafting documentation that must meet strict federal requirements under IDEA (Individuals with Disabilities Education Act), including:

- **Present Levels of Academic Achievement and Functional Performance (PLAAFP)** statements
- **Measurable annual goals** aligned with state standards
- **Specially Designed Instruction (SDI)** specifications
- **Supplementary Aids and Services (SAS)** documentation

The University of Wyoming’s College of Education needed a platform that could help pre-service and in-service educators create compliant, high-quality IEPs faster — without sacrificing the personalized, student-centered approach that effective special education demands.

## Our Solution

### AI Agent Architecture

Fruition designed and deployed a multi-agent AI system where each agent specializes in a specific aspect of IEP creation:

**PLAAFP Writer & Evaluator**  
Generates present level statements based on student data, assessment results, and educator observations. The evaluator agent reviews drafts for compliance with IDEA requirements and suggests improvements.

**Goal Creator & Evaluator**  
Creates measurable annual goals aligned with state standards. The evaluator checks goals against SMART criteria (Specific, Measurable, Achievable, Relevant, Time-bound) and ensures alignment with PLAAFP statements.

**SDI Generator & Evaluator**  
Produces specially designed instruction specifications tailored to each student’s needs and goals. The evaluator validates that SDI directly supports each annual goal.

**SAS Generator & Evaluator**  
Recommends supplementary aids, services, and accommodations based on student needs. The evaluator ensures alignment with goals and PLAAFP data.

**Case Manager**  
Orchestrates the overall IEP creation workflow, maintaining consistency across sections and flagging potential conflicts or gaps.

**Routing Agent**  
Intelligently directs user requests to the appropriate specialist agent based on context and the current stage of IEP development.

## Technical Architecture

- **Frontend**: Next.js 14 + React + TypeScript with real-time collaborative editing  
- **Backend**: NestJS + GraphQL for flexible data querying  
- **Database**: PostgreSQL (Neon) for application data  
- **Cache**: Redis for session management and real-time features  
- **Vector Store**: Pinecone for semantic search across IEP resources and templates  
- **AI**: OpenAI GPT-4 with custom fine-tuning for educational content  
- **Real-time**: WebSockets for collaborative editing and live agent feedback  
- **Infrastructure**: Kubernetes on Digital Ocean with auto-scaling

## Platform Features

- **Real-time collaboration** — multiple educators can work on the same IEP simultaneously
- **Template library** — starting templates based on disability categories and grade levels
- **Compliance checking** — automated validation against IDEA requirements
- **Progress monitoring** — built-in tools for tracking student progress toward IEP goals
- **Version history** — full audit trail of all IEP changes and decisions
- **Data export** — compliant document generation for district reporting systems

## Results & Impact

- **70% reduction** in IEP creation time — from 6+ hours to under 2 hours
- **85% first-pass accuracy** — AI-generated content requires minimal revision
- **100+ concurrent users** supported during peak usage periods
- **99.9% uptime** maintained since deployment
- **6 specialized AI agents** working in coordination for comprehensive IEP support

> “The AI agents understand the nuances of special education documentation in a way I didn’t think was possible. The PLAAFP writer generates present level statements that read like they were written by an experienced special educator.”  
> — Program Coordinator, University of Wyoming

## Security & Compliance

**FERPA Compliance** — the platform implements strict data handling protocols for student education records:

- All student data encrypted at rest and in transit
- Role-based access control with district-level permissions
- Audit logging for every data access and modification
- Data retention policies aligned with federal requirements

**WCAG 2.1 AA Accessibility** — the platform is fully accessible for educators using assistive technologies, with automated accessibility testing in the CI/CD pipeline.

## Scalability

The platform is designed to scale beyond the University of Wyoming to serve school districts and education programs nationwide. The multi-agent architecture allows new specialist agents to be added for additional IEP components without disrupting existing functionality.

> "This platform represents a fundamental shift in how we approach special education planning. The AI agents don't replace educator expertise — they amplify it, handling the tedious documentation while our teachers focus on what matters: the students."  
Dr. Sarah Johnson  
University of Wyoming College of Education