Research shows that the more labelled data you feed an AI model, the smarter it gets.
But in reality, it is only up to a point. Beyond that, every extra labelled example is just cost and effort, quietly changing nothing.
Our researchers patented a technique that precisely locates that point and automatically labels the rest with weak supervision.
It doesn't need more data, just the right amount.
Know more: https://t.co/Kk72EARYTn
#Research #InventingforImpact #ArtificialIntelligence #MachineLearning #DataScience
If a company says its products are "eco-friendly," do you trust it?
As sustainability claims grow, so does the need to verify them.
Our latest research explores an AI model that can accurately identify environmental claims in corporate web content.
Read the research by Diya Saha, Manjira Sinha, and Tirthankar Dasgupta. Green by Design: https://t.co/tUpoO4ozjC
#Research #InventingforImpact #Environment #Sustainability
What if your daily technology could sense when you feel unsupported or disconnected, and adapt to help?
At International Conference on Intelligent User Interfaces (IUI), our researchers introduced SDT-aware neural models that decode human motivation from everyday text.
Read more: https://t.co/4uJr4hVo8T
#Research #InventingforImpact #HumanCentredAI
The next energy frontier calls for innovation on multiple fronts.
At Singapore International Energy Week (SIEW) 2026, Venkatesh Sarangan, Chief Scientist at TCS, outlines a forward-looking, multifaceted approach to address the growing energy challenges related to cost and availability.
From next-generation battery materials to cleaner fuels, TCS Research is advancing innovations for a future-ready energy ecosystem with the right partnerships helping them translating to real-world solutions.
#SIEW2026 #TCS #EnergyInnovation #Sustainability
One global stage. Countless possibilities.
After a landmark 13th season that earned a Guinness World Record, #TCSCodeVita returns with a new challenge: Code Beyond Imagination. Whether you're looking to sharpen your problem-solving skills, gain new perspectives, or connect with a global coding community, TCS CodeVita Season XIV offers an opportunity to learn, grow, and explore what's possible through code.
What awaits?
• Engaging programming challenges
• Opportunities to learn and grow alongside coders from around the world
• Global visibility for your ideas and talent
• Rewards and recognition for standout performances
Register now: https://t.co/ZvpL0I5e0C
The future of energy is taking shape.
At Singapore International Energy Week (SIEW) 2026, Sankaranarayanan Viswanathan, VP and Head of Business Innovation, TCS, puts forward a new vision for how we can deliver greener energy that is economically viable and able to meet the demand of the future.
Watch this video to discover the vision and how this is being made possible.
#SIEW2026 #TCS #FutureofEnergy #EnergyTransition
TCS Research & Innovation(R&I) has recently received a certificate of appreciation at #IITK Samanvay 2026, recognising our contributions to translational research and sustainability initiatives that accelerate India’s innovation ecosystem.
This recognition reinforces the long-standing TCS COIN™ partnership with IIT Kanpur, reflecting our shared commitment in connecting research with real-world applications and to advance purposeful innovation in the future.
@TCS | @IITKanpur
#Research #TCSCOIN #Innovation
Welcoming the next generation of research talents!
We are pleased to introduce Krishnan M, Pavithra K, K. Sai Mineesh Reddy, Rajat Adak and Nishant Gupta from the TCS Research Scholar Program's Cycle 20 cohort.
Here’s to nurturing their research aspirations and supporting their exciting innovations ahead
Know more: https://t.co/LDpXwljNAG
#Research #InventingforImpact #COIN #TCSRSPCycle20 #ResearchScholars
What if every message in Federated Learning was intercepted?
Our researchers, in collaboration with the University of North Carolina at Charlotte, found that sharing model updates rather than complete models can make reconstruction significantly harder for eavesdroppers, offering inherent model protection without sacrificing performance.
Read more: https://t.co/oyrj2ut1eT
#Research #InventingforImpact #FederatedLearning #AIPrivacy
What if the biggest warning sign of a heart attack came before your heart did? Our Cardiac Digital Twin for your heart constantly gets updated with ECGs, wearables, scans and clinical data to help predict risks, personalise treatment and monitor recovery.
Know more: https://t.co/k3Zaudl4HC
#Research #InventingforImpact #Healthcare #DigitalTwin #ArtificialIntelligence
We are proud to be a part of the ServDes India Symposium 2026, hosted at FLAME University Pune on 24th October 2026.
Organized by the Design Research Lab at TCS Research, the symposium brings together service design practitioners, researchers, educators, industry leaders and students to shape the future of services in India.
The day features keynotes, the live jury round of the Student Service Design Competition, where finalist teams present ideas for India’s social sector, and a student poster exhibition.
Together with the ServDes India Chapter, we are helping build a community where research, practice and real-world impact meet.
📍 FLAME University, Pune
📅 24th October 2026 | 9:00 AM onwards
Know more: https://t.co/1yivTyz8f7
#TCSResearch #InventingforImpact #ServDes #ServiceDesign #ServDesIndia
Bringing foundation-model performance to resource-constrained devices remains one of AI's most important challenges.
In healthcare, this challenge becomes even more critical. Accurate skin lesion analysis can support early detection, but many high-performing AI models are too large to run efficiently on mobile and edge devices.
Our researchers developed SparseSegNet, a lightweight, boundary-aware segmentation framework designed for real-time deployment. By combining intelligent sparsity with dual-teacher knowledge distillation, it learns from powerful foundation models while maintaining a compact footprint suitable for edge devices.
Our researchers developed SparseSegNet, a lightweight, boundary-aware segmentation framework designed for real-time deployment.
By combining intelligent sparsity with dual-teacher knowledge distillation, it learns from powerful foundation models while maintaining a compact footprint suitable for edge devices.