Summary
Dheeraj Khanna is a software engineer specializing in computer vision and machine learning with 4+ years of research and industry experience and a MASc in Systems Design Engineering from the University of Waterloo. He has a proven track record of turning research into deployable systems—from a CVPR-workshop runner-up thesis on multi-object tracking that blends Mamba state-space models with self-attention to productionized pipelines for high-throughput industrial part tracking. His work spans sensor fusion, multi-sensor calibration, real-time deployments on edge devices, and large-scale annotation automation achieving annotation speeds of 10 FPS. Comfortable in Python, C++, PyTorch, OpenCV and Docker in Linux environments, he has migrated critical pipelines to Libtorch for low-latency production use and optimized perception stacks on embedded GPUs. Now at Mecka AI, he’s pursuing full-time roles where applied ML and CV solve meaningful real-world problems while remaining open to collaborative research opportunities. An engineer who enjoys bridging theoretical depth with pragmatic engineering, he often focuses on making complex models robust and efficient in constrained deployment settings.
11 years of coding experience