Upload once
Upload the answer key and scanned student papers for a class, subject and exam.
IgniteXVIDHYASETU / EXAMINATION WORKSPACE
Every paper.
Checked with care.
VidhyaSetu gives schools an AI-assisted first evaluation of handwritten answer sheets, while teachers remain responsible for reviewing, correcting and approving final marks.
Answer-key alignedTeacher approvedBatch processing
01 / WHAT CHANGES
AI does the repetitive first pass. Teachers keep the decision.Evaluate question by question
VidhyaSetu reads handwritten answers and suggests question-wise marks with visible reasoning.
Review where it matters
Teachers focus on corrections and approval instead of repeating the full checking process.
Move to results faster
Approved marks flow into class-level results so schools can prepare outcomes sooner.
02 / INSIDE THE PRODUCT
Built around the school examination workflow.TEACHER REVIEW
See the paper and AI suggestion together.
Review extracted answers, question-wise marks and reasoning before approving or correcting the score.
EXAM WORKSPACE
One place for batches, review and results.
Support teams process papers while teachers work through review queues and results.
03 / MILESTONES ACHIEVED
Capabilities already demonstrated in the current product.End-to-end evaluation
Answer-key upload, handwritten answer-sheet processing, question-wise scoring and teacher review are working as one flow.
Teacher approval
Teachers can inspect AI suggestions, correct marks and approve the final result.
Batch processing
ZIP-based batch processing and automatic retry handling reduce manual intervention.
School workflows
Admin, teacher and support roles cover setup, exams, processing, review queues and results.
04 / PERFORMANCE MATRIX
Measured internal benchmark; actual production performance varies by workload.87 sec3 papers / 6 pages
Internal NVIDIA L4 batch benchmark.
≈29 secper paper in that test
Observed pace for the tested paper mix and configuration.
100/hroperational target
Target for ongoing GPU and workflow optimisation.
Throughput depends on page count, scan quality, handwriting, question type, model choice, GPU capacity and the amount of teacher review required.
05 / TIME & COST BENEFIT
Turn full-paper checking into focused review.Return teacher time
to teaching.
If a teacher currently spends 5 minutes checking one paper, 300 papers represent about 25 teacher-hours. If AI performs the first evaluation and teacher review averages 1 minute per paper, the same batch would require about 5 teacher-hours of review.
That example returns roughly 20 teacher-hours for one 300-paper batch. The exact saving depends on your current checking time, review policy and paper complexity.
25 hmanual example
300 papers × 5 minutes each.
5 hreview example
300 papers × 1 minute teacher review.
20 hteacher time returned
Illustrative saving for that example batch.
SEE IT WITH YOUR OWN PAPER
Bring one answer key.
We’ll show the workflow.
A focused demo can use a representative school paper so your team can judge the review flow and speed directly.