KOLKATA: IIT Kharagpur will use artificial intelligence to identify early indicators of academic stress among students, such as a sudden drop in CGPA (cumulative grade point average), accumulation of uncleared papers and a sustained decline in continuous performance assessment, enabling the authorities to intervene in time and ensure their mental well-being.The institute, which witnessed five suicides in 2025 and two in April 2026, has strengthened its faculty mentorship programme to support students in crisis. It will also award one extra academic credit to hostel wing representatives who perform their responsibilities effectively.
The hostel wing representatives, enrolled under the ‘Saarthi’ programme, are required to spot behavioural changes in their peers and alert the authorities if any of them needs counselling.The tech school has also decided to conduct supplementary exams twice a year so students don’t have to wait for long to clear backlogs.“Academic stress should not become invisible until it is too late,” IIT Kharagpur director Suman Chakraborty said. “We want to move from crisis response to early support. By combining data-driven alerts… faculty mentorship, accessible academic recovery mechanisms and professional well-being support, we are building a system that identifies concerns early and responds with compassion.”AI to send alert on CGPA drop, backlog | page 4The AI-enabled risk-flagging mechanism will provide alerts on sudden CGPA drop, accumulation of backlogs, repeated absence from class and a sustained decline in continuous assessment performance to faculty advisers.An official said, “Faculty advisers often say since they have many students under their supervision, it becomes difficult to spot the critical cases. So, if cases are automatically highlighted, it will be easier for them to provide immediate support.”The institute is set to roll out a faculty-led early detection mechanism, known as ‘Mitra’, for UG students as a pilot, in which a teacher or mentor will be assigned 10 to 12 students. “Mentors must meet students once a month and log their mood, stress and isolation signals in the Setu app. A case, if flagged, will trigger an alert to the counselling centre within 24 hours,” said an official.Director Chakraborty said, “We have re-structured the hostel wing representative system: students who will perform their task effectively as wing representatives will get one extra academic credit. There will also be incentives for the best wing representative.”
