PITCH Accepted to AsiaCCS 2025
We can now fight off deepfake audio calls by outsmarting them using PITCH. Check it out here.
I am a Senior Machine Learning Engineer and researcher specializing in production-ready AI systems with expertise in computer vision, generative AI, and security applications. Currently completing my Ph.D. in Computer Science at NYU Tandon, where I work with Professors Nasir Memon and Chinmay Hegde. My technical foundation includes a dual degree in Computer Science and Mathematics from BITS Pilani, India, complemented by extensive experience in architecting, optimizing, and deploying ML systems at scale.
My work focuses on building high-performance ML architectures for content verification, media security, and real-time analytics. I've engineered fault-tolerant ML pipelines processing 18+ TB of streaming data with 99.8% uptime, developed award-winning diffusion models for identity-preserving image generation (Best Paper Award at IJCB 2023), and created novel deepfake detection frameworks with 88% accuracy (featured in WIRED). My technical expertise spans transformer architectures, diffusion models, distributed ML systems, MLOps, computer vision, speech processing, and optimization techniques for production environments.
I've published at top ML and security venues including NeurIPS, ICLR, IEEE Security, and AsiaCCS, with my research on multimodal deepfake detection attracting $100,000 in industry funding. My approach combines rigorous ML engineering with practical solutions that bridge the gap between cutting-edge research and production-ready applications. I've led teams of 6+ engineers through the complete ML lifecycle, implemented CI/CD for ML systems, and reduced model training time by 70% while improving accuracy through advanced optimization techniques.
I am currently seeking ML Engineering and Applied AI roles where I can leverage my expertise in building large-scale ML systems, generative AI, computer vision, speech processing, and security to deliver impactful solutions with measurable business value. I'm particularly interested in positions involving real-time ML systems, content verification, generative AI applications, and MLOps at scale. Please explore my publications and projects to see examples of my work.
We can now fight off deepfake audio calls by outsmarting them using PITCH. Check it out here.
We can now capture data off of radio live streams using WavePulse. Check it out here.
I fact-checked an audio deepfake at Reuters. Read the article here.
Our work on GOTCHA, a real-time deepfake detection system, was featured in WIRED magazine! The article discusses how GOTCHA uses challenge-response verificat...