Projects

My research sits at the intersection of algorithms and society: building systems that are not just efficient, but fair, transparent, and accountable to the people they affect. This translates into work at several scales: leading a dedicated research group that investigates bias and fairness in online platforms; an international partnership that extends this agenda to the next generation of decentralized platforms; and two national-scale infrastructure efforts that address a more foundational problem for Indian AI — the scarcity of high-quality, India-representative data and open learning resources. Together, these projects reflect a common thread: using rigorous, data-driven research to make algorithmic systems work more equitably, and to strengthen the data and knowledge infrastructure that India's AI ecosystem depends on.

Head, Max Planck Partner Group on Responsible AI

I lead the Partner Group of the Max Planck Institute for Software Systems (MPI-SWS), building on a decade-long collaboration with Prof. Krishna Gummadi. The partner group focuses on Responsible AI for Decentralized Online Platforms, studing how fairness, accountability, and transparency can be embedded directly into emerging decentralized systems such as BlueSky (social media), ONDC (e-commerce), and UPI (payments). We aim to move beyond diagnosing bias in traditional centralized platforms to building equitable infrastructure from the ground up for decentralized ecosystems, including fair content moderation, cooperative gig-work marketplaces, and decentralized governance frameworks

Head, REsponsible AI Lab (REAL)

My research group at IIT Kharagpur, where we study fairness and bias in algorithmic decision-making across online platforms, from e-commerce and social media to gig-work and recommendation systems. The lab combines large-scale data analysis with algorithmic interventions, drawing on machine learning, computational social choice, and economic theories, to understand where algorithms cause harm and to design mechanisms that make these systems fairer for all stakeholders.

Website

PI, Data Hub KGP

A centralized repository of high-quality, curated, and annotated datasets built for India's unique linguistic, cultural, and socio-economic landscape. Since most AI training data originates from Western contexts, Data Hub KGP fills this gap with India-representative datasets spanning governance, education, law, agriculture, tourism, low-resource Indian languages, renewable energy, responsible AI, and smart cities — powering AI models that are genuinely built for India.

Website

Co-PI, National Digital Library of India (NDLI)

India's largest virtual repository of learning resources, sponsored by the Ministry of Education, and operated at IIT Kharagpur. NDLI goes beyond a searchable repository to offer user-group-specific services for school and college students, researchers, and job aspirants, with content and interface support across most major Indian languages — democratizing access to educational resources at a national scale.

Website