scientist / researcher / engineer
Ris(h)i Jaiswal
Principal ML Scientist at Ludwig Computing | PhD, Purdue University
Advancing efficient LLM inference through probabilistic computing. Bringing physics-inspired algorithms and hardware together.
Research Interests
Efficient LLM Inference
Probabilistic computing for efficient large language model inference. Exploring the intersection of sampling, machine learning, and computing systems.
Bio-Inspired Learning
Energy-efficient neural architectures inspired by biological systems. Alternatives to backpropagation, probabilistic computing, and stochastic neural networks.
Physics-Inspired AI
Embedding physics into learning and inference. Probabilistic graphical models, Boltzmann machines, quantum-inspired optimization, and physics-informed learning.
Hardware & Systems
GPU kernels, FPGA prototyping, and hardware-algorithm co-design. Bridging algorithms and silicon for petaflops-scale probabilistic computing.
Journey
Principal ML Scientist
Ludwig ComputingProbabilistic computing for efficient large language model inference.
Cofounder & CTO
Transio AISpatial intelligence and world models for AI systems.
PhD, Electrical & Computer Engineering
Purdue UniversityEnergy-efficient deep learning, probabilistic computing, quantum-inspired optimization. Advisors: Supriyo Datta & Joseph G. Makin.
Research Intern
MicrosoftFormal verification and assertion-based validation of custom AI accelerator architectures.
Senior Design Engineer
NXP SemiconductorsVerification of DSP, RADAR-based ADAS subsystems, and neural network accelerators. US Patent holder.