Manual Remediation vs. Aelira: Planning Costs and Review
Compare manual repair with supported AI-assisted fixes, human review, and an agreed processing allowance.
The Reality of Manual Remediation
Planning Manual Review at Scale
Check the date that applies to your institution
To manually fix 10 courses (50 hrs each)
Review capacity for larger inventories
Estimate inventory, staff capacity, supported automation, and human review against your applicable institutional milestone.
Manual Remediation
Reality: Hiring staff or faculty to manually fix accessibility issues
- Full control over every fix
- Works for small-scale projects (1-5 courses)
- Illustrative assumption: 40-80 hours per course; replace with measured effort
- $3,000-$6,000 per course under the stated illustrative assumptions
- Inconsistent quality (depends on skill level)
- LaTeX conversion requires source, equation, and assistive technology checks
- Faculty burnout (no compensation for extra work)
Aelira (AI-Assisted)
Reality: Supported partial fixes with human review and unresolved manual work
- Measure processing, review, and manual repair time per course
- $1,499/mo defined processing allowance
- Batch processing for supported files within the agreed allowance
- Output quality needs human checks
- Supported LaTeX/MathML conversion needs review
- Measure faculty and reviewer time in a pilot
- Scalable to entire departments
Planning Assumptions to Test: Manual vs. Aelira
Scroll horizontally if needed to view every column.
| Feature | Manual Time | Manual Cost | Aelira Pilot Measurement |
|---|---|---|---|
| PDF Remediation (50 pages) OCR, alt text, heading structure, contrast fixes | 10-15 hours Per document | $750-$1,125 | Measure OCR, review, and remaining fixes |
| PowerPoint Alt Text (100 images) Write contextual descriptions for each image | 8-12 hours Per presentation | $600-$900 | Measure Image-context review and corrections |
| LaTeX MathML Conversion (50 equations) Convert math equations to accessible format | 15-20 hours Extremely tedious | $1,125-$1,500 | Measure Conversion and math review |
| Video Caption Cleanup (10 videos, 1hr each) Fix auto-caption errors, technical terms | 10-15 hours Tedious work | $750-$1,125 | Measure Caption editing and quality checks |
| TOTAL (Illustrative Task Set) | |||
43-62 hours | $3,225-$4,650 | Pilot result needed No fixed time saving is established | |
Plan Review Capacity from Your Inventory
Inventory
Count file types, pages, equations, and videos. Include new uploads, inaccessible publisher materials, and files that cannot produce a usable artifact.
Measure
Time a representative sample through scanning, supported fixes, saved-output review, and manual repairs. Aelira throughput and time savings are not established by this page.
Schedule
Allocate reviewers and content owners, then estimate completion using measured hours and available staff time. US covered public entities generally have 2027 or 2028 dates by population and category; Australian DDA/DSE obligations have no matching US deadline.
LaTeX: Source and Equation Review
Why LaTeX Conversion Is So Hard
- Complex syntax - Nested commands, macros, custom packages
- No visual editor - Choose a conversion path suited to the source and readers
- Thousands of equations - Physics course = 200+ equations
- Error-prone - One typo breaks the entire equation
LaTeX: \int_0^\infty e^{-x^2} dx = \frac{\sqrt{\pi}}{2}
↓ Manual MathML conversion (30+ lines of XML)
What to Evaluate in Aelira
- Automated parsing - Test supported amsmath and physics-package cases
- Bulk processing - Measure supported conversion and equation review
- ARIA labels - Generated descriptions require a mathematics review
- Validation - Syntax checks do not establish semantic or accessibility correctness
Illustrative Labor Inputs (per STEM course)
Manual: 15-20 hours × $75/hr = $1,125-$1,500
Aelira: processing and review time must be measured; allowance agreed for $1,499/mo
A typical STEM faculty concern: a single quantum mechanics course can carry 300+ LaTeX equations. Converting them to accessible MathML by hand can take weeks, and most professors are subject-matter experts, not web developers. They need tooling that handles the conversion for them.
When Manual Remediation Makes Sense
When to Go Manual
- Very small scale (1-5 courses only)
- High-touch content (complex graphics requiring expert review)
- Legal review required (medical, legal fields)
- One-off projects (single PDF, not ongoing)
- Learning exercise (training staff on accessibility)
When to Evaluate a Combined Workflow
- Departments with 10+ courses (scale problem)
- STEM content with LaTeX (source and math review needed)
- Time constraints depend on your institution's obligations and inventory
- Limited budget (manual = $3K-$6K per course)
- Faculty without compensation (burnout risk)
Evaluate a Hybrid Approach
Evaluate supported Aelira batch workflows (PDFs, PowerPoints, LaTeX, videos), with human review of every saved candidate and manual repairs where needed (complex diagrams, specialized graphics).
Use automation for supported partial fixes, then budget for expert review and unresolved manual work. Measure both on a pilot corpus.
Build a Cost Model with Measured Inputs
Manual cost = measured repair and review hours × your staff rate. AI-assisted cost = processing and review labor + unresolved manual repairs + subscription + deployment costs.
A six-month Department subscription at $1,499/mo is $8,994 before additional labor, infrastructure, and any allowance adjustments. This is a price calculation, not a measured saving or a conformance result.
The difference can be positive, zero, or negative. Keep each amount in the same currency, and include review of the saved output even where a scan score improves.
Explore your own cost assumptions in the ROI calculatorMeasure Processing, Review and Manual Work
Discuss a scoped pilot and measure total review, correction and manual work against your baseline. Confirm the duration and terms in writing.
Join a bounded pilot to test supported workflows against your institution's obligations.
Comparisons are based on publicly available information as of August 2026 and reflect Aelira's views. Competitor features and pricing change over time; verify current details with each vendor. This page was authored by Aelira.