Proudly based in Williamsburg. Brooklyn, NY.
Forecasting as a business function involves the synthesis of objective and subjective inputs. Models learn to pull forward patterns from objective history into an estimate of the future. The best models have always included subjective context in the form of feature engineering, choice of sampling weights, selections of training sets, well chosen priors, and so on.
Chronulus AI is the first company to provide a purpose built forecasting and prediction platform that eliminates the bottle neck of manually injecting subjective inputs into the forecast model lifecycle. By eliminating this pain point, we are able to generalize forecasting and prediction across a broad spectrum of use cases that were previously out of reach for most businesses.
About the Founder
Chronulus AI Inc. was founded in 2024 by Jeremy Oldfather.
Jeremy is a Deep Learning Scientist who specializes in Forecasting. He has accumulated 20 years of experience across research, engineering, and machine learning solution delivery, including delivery of solutions in 14 programming languages.
Over his career, he has built production deep learning systems for numerous fortune 500 companies including Novartis, Nationwide Insurance, Best Western, BP, and Bridgestone Tire. Before launching Chronulus AI, he rebuilt Amazon Retail's core forecasting package (deepts), modernizing it from mxnet to pytorch. Today, deepts is used by Amazon SCOT to forecast unconstrained demand for hundreds of millions of products worldwide on a daily basis.
Jeremy also worked for 5 years in the Division of Research and Statistics at the Federal Reserve Board of Governors, contributing to areas including macroeconomic forecasting, quantitative surveillance, and financial stability in the wake of the Great Recession.
Jeremy holds an MS in Mathematics & Statistics from Georgetown University that he earned while working full-time at the FRB. He also holds a triple major in Mathematics, Economics, and Russian Language from Indiana University-Bloomington.
When Jeremy isn't back-solving to perfect Chronulus agents, you will mostly like find him at Society Billiard back-solving easier problems.
Back-solving with Michael Secondo at Society Billiards