Safety Gaps in Health: IIT Students Propose AI-Driven Solutions
To address this, the team proposed the Composite Weighted Mass Balance (CWMB) Framework.
Hyderabad:Students from IITs and other engineering institutes have proposed AI-driven solutions to address long-standing regulatory and safety gaps in drug development and monitoring under an initiative of Novartis India’s NEST 2.0 (Nurturing Excellence, Strengthening Talent).
This initiative is led by the Novartis Development Hub in India.
The participating teams included NovaNexus from IIT Patna, Parzival Prime from Gokaraju Rangaraju Institute of Engineering and Technology and IIT Bombay, and Codeies from the Vishwakarma Institute of Technology, Pune.
Sadhna Joglekar, head of the Development Hub, India, said, “We were encouraged by how naturally these young minds integrated technology and AI into their solutions. The focus now is on evaluating whether these ideas have the potential to deliver meaningful and scalable impact for the real world.”
He said, “Team Codeies addressed a major challenge in pharmacovigilance — incomplete reporting of drug side effects. When patients report adverse reactions, pharmaceutical companies must collect detailed information to assess risk and meet global safety regulations. However, initial reports are often incomplete.”
Current follow-up methods rely on repeated phone calls, emails and long questionnaires. These approaches frequently fail because doctors and patients are busy, people become fatigued with repeated follow-ups, language barriers exist, and there is “growing mistrust due to spam or scam communications.”
To solve this, he said, the team developed SmartFU, an AI-driven digital follow-up platform. SmartFU automatically identifies missing safety information in reports and generates only relevant, critical follow-up questions instead of lengthy generic forms. “It prioritizes high-risk cases, adjusts communication based on whether the reporter is a doctor or patient, and selects the most suitable communication channel such as email or SMS.”
Team NovaNexus focused on challenges in forced degradation testing, where medicines are exposed to heat, light and humidity to study how they break down. Scientists calculate mass balance to check whether the total drug and its breakdown products add up correctly.
The difficulty, he said, is that analytical instruments cannot always detect every breakdown product. Some substances may be below detection limits, and detector responses can vary. When results do not add up neatly, companies struggle to determine whether it is a real chemical problem or simply a limitation of the testing method, leading to delays and repeated experiments.
To address this, the team proposed the Composite Weighted Mass Balance (CWMB) Framework.
“Instead of trying to artificially correct missing values, the framework evaluates how reliable the testing method is before interpreting results. It assigns confidence weights based on assay reliability, detection limits, chromatographic performance, and detector response differences.”
