Valery Manokhin, PhD, is an independent machine learning researcher, author, and educator specialising in conformal prediction, uncertainty quantification, and modern time series forecasting. He holds a PhD in Machine Learning from Royal Holloway, University of London, where he was supervised by Professor Vladimir Vovk — the creator of conformal prediction. He also holds an MBA from the University of Warwick, an MSc in Computational Statistics and Machine Learning from University College London, and a Certificate in Quantitative Finance (CQF).
His doctoral research produced the first multi-class Venn-Abers predictor (PKPD — Pairwise Kriging Probability Distribution), extending distribution-free calibration guarantees beyond binary classification. His 72-dataset empirical study remains one of the largest systematic evaluations of classifier calibration ever published.
Valery’s books have reached readers across 100+ countries. His titles include Mastering Modern Time Series Forecasting, Applied Conformal Prediction, Mastering CatBoost, and the English translation of Kiselev’s classic Arithmetic — now available on Amazon KDP. He publishes through Gumroad, Amazon, and Leanpub.
He maintains the widely cited Awesome Conformal Prediction repository on GitHub and has built a technical community of 100K+ followers on LinkedIn and Twitter/X. His work has been read by practitioners and researchers at organisations including Amazon, Deloitte, Microsoft, EY, and JPMorgan.
He writes for people who already understand the fundamentals and demand rigour, mathematical precision, and production-grade thinking — not hype.
