Akshay Singh

I am a PhD Graduate from the Department of Computer Science and Engineering at the Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, under the supervision of Prof. Rahul Thakur.

My research interests lie at the intersection of federated and personalized learning, edge AI, and resource-efficient deployment across real-world, privacy-sensitive environments. I am particularly driven by challenges involving data heterogeneity, limited compute, and privacy preservation, especially in applications like multilingual NLP, intelligent transportation, and human-centered adaptive systems. Beyond algorithmic development, I am deeply passionate about building end-to-end AI solutions on resource-constrained platforms such as Arduino, Raspberry Pi, and NVIDIA Jetson, with a focus on microcontroller-level deployment. This includes optimizing model footprint and inference pipelines to run efficiently on-device. My PhD work has been rooted in this philosophy—developing FL-based abuse detection on systems for low-resource Indian languages and sensor-driven driver activity models, leading to practical, scalable systems that reflect both academic rigor and real-world viability. Going forward, I am excited to contribute to teams working on embedded AI, edge ML, and federated systems where meaningful deployment meets cu ng-edge research.