Skip to content
Splashgain

How marking works

From scanned script to reviewed result

Secure scanning & upload

Physical answer sheets are scanned and uploaded securely — the digital workflow removes shipping, storage and examiner travel from the process entirely.

    AI text extraction & evaluation

    An OCR engine converts handwriting to digital text, then a domain-tuned model compares each answer to the model answer and explains its mark adjustments.

      Faculty review & reports

      Faculty review AI evaluations, add feedback, and generate progress, performance and question-level analytics before results are published.

        Identity masking & re-evaluation

        QR code identity masking supports blind grading, and an automated re-evaluation workflow tracks escalations and re-checks against a timeline.

          Role-based access control

          Admin, COE, examiner, scanner, moderator and re-evaluator roles keep the marking process auditable and aligned to institutional controls.

            Four evaluation modes

            Default, easy, moderate and strict marking modes let an institution match AI grading rigor to the exam’s stakes.

              The transformation

              45 days of grading, down to 8

              • 8 days

                Result processing time (down from 45 days)

              • 80%

                Reduction in operational cost per evaluation

              • 200+

                Faculty hours saved per evaluation cycle

              • 5M+

                Evaluations completed

              FAQ

              Questions about onscreen marking

              How much faster is AI onscreen marking than manual grading?

              Institutions typically move from a 45-day result cycle to 8 days — an 82% reduction — while cutting operational cost per evaluation by roughly 80%.

              Does AI replace human evaluators?

              No — it works alongside them. Faculty review the AI’s scoring and feedback on every answer sheet, add their own comments, and sign off before results are finalized.

              What languages does onscreen marking support?

              English, Hindi, Marathi, Tamil and Telugu are supported for evaluation, with AI feedback translatable for review.

              Is Eklavvya’s onscreen marking patented?

              Yes — Eklavvya holds an India and Worldwide PCT patent for its Onscreen Marking System.

              How does identity masking work?

              A QR code conceals the student’s identity on the digitized answer sheet to support blind grading, and an automated re-evaluation workflow tracks any escalated re-checks against a timeline.

              See onscreen marking on your own answer scripts.

              Bring a sample batch — we'll show you AI evaluation alongside your evaluators.