Summary
Arjun Pankajakshan is a Machine Learning Research Engineer with a PhD in Computational Audio Analysis and eight years of experience building state-of-the-art machine listening systems. His research and applied work span sound event detection, audio tagging, sequence modelling and attention-based transformers, with practical deployments including real-time bird detection and audio deepfake detection. At Queen Mary University of London he advanced CRNN and time-restricted self-attention models and integrated source-separation into multi-head attention, improving performance on UrbanSED and DCASE benchmarks. He is fluent in Python, PyTorch/TensorFlow, Librosa and model deployment tooling (Hugging Face, LangChain, Streamlit), and has hands-on experience scaling multi-class bird-species classifiers to 200 classes. Based in London, he blends rigorous academic research with product-focused engineering at RediMinds, often synthesizing datasets (Scaper) and leveraging pre-trained audio backbones to accelerate development. A less obvious strength is his cross-domain fluency—from TTS and NLP experiments to classical signal-processing techniques—enabling creative, hybrid solutions for complex audio tasks.
8 years of coding experience
2 years of employment as a software developer
MATRICULATION Science, MATRICULATION Science at SNHSS, SREEKANDESWARAM ALLEPPY
Engineer’s Degree Electronics and Communications Engineering, Engineer’s Degree Electronics and Communications Engineering at Govt. Engineering College, Sreekrishnapuram
Doctor of Philosophy - PhD Computational audio analysis Deep learning, Doctor of Philosophy - PhD Computational audio analysis Deep learning at Queen Mary University of London
Master of Technology (MTech) Communication Engineering & Signal Processing, Master of Technology (MTech) Communication Engineering & Signal Processing at GEC T