Self-hosting is live: github.com/Aelira-AI/aelira-core
Aelira is open source (MIT + AGPL), released as v0.9.5 stable. The deployment guides in the repository are the authoritative self-hosting reference; the pages here are a summary. Operational hardening is ongoing and tracked in public issues.
Getting Started with Aelira
Aelira is an AI-powered accessibility compliance platform that helps universities and organizations meet WCAG 2.1 standards. Scan documents, generate fixes, and ensure compliance before the April 2027 deadline.
What is Aelira?
Aelira is a comprehensive accessibility compliance platform designed specifically for higher education institutions facing the April 2027 DOJ ADA Title II deadline. With the new regulations requiring all digital content to meet WCAG 2.1 Level AA standards, universities need a scalable solution that can process thousands of documents, presentations, videos, and course materials without overwhelming faculty and staff.
Unlike traditional accessibility scanning tools that simply identify issues, Aelira uses AI-powered remediation to automatically generate working code fixes for accessibility violations. Our platform combines advanced document processing (OCR, structure tagging, contrast analysis) with state-of-the-art language models to produce standards-compliant content that actually works for users with disabilities.
The platform supports the full range of educational content formats: PDF documents with OCR and structure tagging, PowerPoint presentations with automated alt text and contrast checking, LaTeX mathematical notation converted to accessible MathML, video and audio files with transcription and caption generation, Excel spreadsheets with proper data table markup, and Word documents with heading structure analysis.
Aelira integrates directly with the tools universities already use, including Canvas LMS, Blackboard Learn, Google Workspace, and Microsoft 365, to automatically detect new content, scan for accessibility issues, remediate problems, and sync fixed files back to the cloud. Faculty don't need to change their workflows; Aelira works in the background to ensure compliance.
Why Aelira?
Deep AI-Generated Remediation
Tools like YuJa Panorama and Anthology Ally remediate common issues, but stop short of deep tag-tree, reading-order, table, and native LaTeX source remediation. Aelira generates working code, proper alt text, corrected color contrasts, and accessible markup, saving faculty 40-80 hours per course.
Complete Platform Coverage
Aelira works with Google Workspace (60% market share), Microsoft 365 (40%), Canvas (35%), Blackboard (25%), Moodle, and Brightspace. One platform covers your entire tech stack.
LaTeX/MathML Support for STEM
95-99% of mathematicians use LaTeX, yet few accessibility tools remediate it natively. Aelira converts LaTeX equations to accessible MathML with proper ARIA labels, supporting advanced notation like ChemFig, mhchem, physics, and TikZ diagrams.
Bulk Processing & Automation
Process entire course directories with CLI batch commands. Real-time webhooks detect new files in Google Drive or OneDrive and automatically scan them. No manual uploading required.
Privacy-First Architecture
Self-hosted deployments can run open-weight inference entirely on institution-controlled infrastructure. Hosted routes are region- and provider-specific, with FERPA responsibilities, retention, deletion, and subprocessors documented in the institution agreement.
90% More Affordable
Enterprise tools like YuJa Panorama use custom, quote-based pricing (not publicly listed). Aelira publishes a $1,499/month managed department starting price with a defined Processing Unit allowance, plus clear self-hosted and institution deployment paths.
Quick Links
What Aelira Does
Common Use Cases
STEM Faculty: LaTeX Math Accessibility
A physics professor has 200 problem sets with LaTeX equations. Each equation needs to be converted to accessible MathML with ARIA labels. Manual conversion would take 80+ hours. With Aelira's LaTeX processor, they scan the entire directory and receive MathML-converted files in minutes.
Time saved: 80 hours → 10 minutes | Cost: $6,000 in labor → $1,499/mo subscription
Online Course Providers: Bulk Video Captioning
A university has 500 lecture videos without captions. Auto-captions from YouTube are unacceptable per ADA regulations. Aelira's transcription engine generates accurate VTT/SRT captions using open-source transcription, with speaker diarization and timestamp correction.
Time saved: 500 hours → 50 hours (review) | Cost: $37,500 manual captioning → $1,499/mo
Instructional Designers: PowerPoint Remediation
An instructional design team manages 1,000+ PowerPoint presentations across 50 courses. Each deck needs alt text for images, contrast fixes for text, and proper reading order. Aelira's PowerPoint scanner processes all files in bulk, generates alt text with an open-source vision model, fixes low-contrast text, and validates structure.
Time saved: 200 hours → 20 hours (review) | Cost: $15,000 in labor → $1,499/mo
IT Departments: Campus-Wide PDF Compliance
A university IT department discovers 10,000 PDFs on their website lack structure tags and alt text. Aelira's PDF remediation engine uses OCRmyPDF to add structure tags, detects images requiring alt text, and generates compliance reports showing progress toward the April 2027 deadline.
Time saved: 1,000+ hours → 100 hours (review) | Cost: Avoiding ADA lawsuits: priceless
Frequently Asked Questions
Is Aelira open source?
Yes — live today at github.com/Aelira-AI/aelira-core. The v0.9.5 stable release is the complete platform — scanning, remediation, LMS integrations, the React dashboard, and the CLI — with nothing feature-gated. The engine is AGPL-3.0; the command-line client is MIT so you can embed and script against it freely. Self-host all of it on your own infrastructure for free. See our open source page for details.
How accurate is the AI alt text generation?
Aelira uses an open-source vision-language model optimized for accessibility descriptions. Accuracy is 85-90% for common educational content (charts, diagrams, photos). We recommend human review for complex scientific imagery, but AI alt text is dramatically better than no alt text, and far faster than manual writing.
Does Aelira work with Canvas/Blackboard?
Yes for Canvas LMS: a native LTI 1.3 integration, verified end to end against a live Canvas instance. Install the Aelira app, and it will automatically scan course files, detect accessibility issues, and provide one-click remediation, all within your LMS interface. D2L Brightspace is in beta; Blackboard Learn and Moodle are experimental.
Can I process files in bulk?
Absolutely. Aelira's CLI tool supports bulk directory scanning: aelira scan pdf ./courses/. You can also set up real-time webhooks to automatically scan new files uploaded to Google Drive or OneDrive. See our feature documentation for details.
How much does Aelira cost for universities?
Managed department service starts at $1,499/month per department with an order-form Processing Unit allowance and priority support. Institution plans cover supported self-hosting/BYOC, pooled cloud, dedicated cloud, and high availability. See our pricing page for full details and ROI calculator.
What about data privacy and FERPA compliance?
All AI processing runs locally on your infrastructure using self-hosted open-source models via Ollama. Your self-hosted deployments can keep educational materials on institution-controlled infrastructure. Hosted routes are separately disclosed by provider, region, retention, and subprocessor. FERPA responsibilities are documented in the institution agreement; no infrastructure choice alone guarantees compliance. See our privacy policy for details.
When is the April 2027 ADA Title II deadline?
The DOJ ADA Title II rule requires all public universities to meet WCAG 2.1 Level AA standards by April 26, 2027 (large universities) or April 26, 2028 (smaller universities). These dates were extended by one year under the DOJ Interim Final Rule published April 20, 2026; the rule's substance is unchanged. This includes all digital content: websites, course materials, PDFs, videos, presentations, and learning management systems. Non-compliance can result in lawsuits, OCR investigations, and loss of federal funding. See our higher education page for details.
Get Started in 3 Steps
Clone the repository
git clone https://github.com/Aelira-AI/aelira-core.gitStart the services
cd aelira-core && docker compose up -dOpen the dashboard
Open http://localhost:5173 in your browserNeed Help?
Join our community or reach out for support.