Summary
Rehan Ahmad is a senior machine learning engineer with 10 years’ experience applying speech, NLP and computer vision research to production-grade systems across academia and industry. He holds a PhD in multimodal speaker diarization and completed a postdoc focused on unsupervised domain adaptation for ASR, bringing deep expertise in multilingual and self-supervised speech modeling. Rehan has led development of conversational AI agents, persona-based voice agents, TTS fine-tuning (StyleTTS2, Orpheus) and avatar lipsync across multiple languages, and has productionized pipelines using Argo, MLflow, Docker/Kubernetes and AWS. His background spans FPGA-based real-time video processing to cutting-edge speech-quality estimation, reflecting a rare blend of hardware, signal processing and modern deep learning skills. Based in Islamabad, he combines research rigor with hands-on engineering—often deploying end-to-end ML systems that bridge academic advances and practical voice-agent products.
9 years of coding experience
10 years of employment as a software developer
MS Computer Engineering, MS Computer Engineering at National University of Sciences and Technology (NUST)
Doctor of Philosophy - PhD(EE) Speech Processing | Multimodal Speaker Diarization, Doctor of Philosophy - PhD(EE) Speech Processing | Multimodal Speaker Diarization at International Islamic University, Islamabad