Empowering tool: MCC’s AI voice scribe to write exams for blind students

CHENNAI: Visually challenged students at Madras Christian College will no longer have to depend on manual scribes to appear for exams, as the institution is set to introduce an indigenous AI-powered ‘Voice Chat Scribe’ software to help visually challenged students independently write examinations.

The software has been developed in-house with the help of the MCC-MRF Innovation park, Tambaram.

Principal Prof Dr P Wilson said the number of volunteers available for scribing had been declining and that manual scribes did not always accurately reflect the articulation of students.

“MCC has been making efforts to make higher education inclusive and equitable for visually impaired students through its Student Services for the Differently Abled ecosystem and skill development initiatives with the help of the Help the Blind Foundation. This is another inclusive feature added to its services through an AI-powered interactive digital scribing platform with a voice-to-text conversion engine developed indigenously with higher accuracy by MCC-MRF Innovation Park student interns,” he said.

The new system is expected to be introduced for the third internal continuous assessment before the November 2026 semester exams.

Madras Christian College said the existing manual scribe system had several limitations, including transcription errors and in some case assistants could unintentionally influence students while writing their answers.

The new system will allow students to dictate their answers in English directly into a secure platform without depending on another person to transcript their thoughts.

The AI scribe will enable students to go through questions, repeat audio prompts and check word counts through voice commands or keyboard shortcuts. It will also support formatting functions such as bolding, underlining, rearranging paragraphs and inserting sub headings through spoken commands.

The command and dictation mechanism has been incorporated to distinguish software instructions from the candidate's answers. The system is designed to provide high accuracy voice recognition for both English and Tamil.

For examination security, the software will operate on a restricted and confidential server environment to prevent malpractice and protect assessment integrity.

Following discussions involving the Office of the Controller of Examinations, deans, heads of departments and the student support for differently abled team, the college decided to implement the system.

Training sessions and dry runs will be conducted for visually challenged students and departmental nodal officers before the formal assessments scheduled for next month.