Neuralk is a predictive AI company focused on helping enterprises make predictions from structured, tabular data. Its pretrained foundation model is designed to work with rows and columns without requiring users to train a model, build an machine-learning pipeline, or manage infrastructure. Customers send data through an API and receive predictions, with support for classification and regression, mixed data types, and datasets of up to 50 million rows.

The platform serves use cases including fraud detection, demand and sales forecasting, product classification, credit scoring, predictive maintenance, customer churn, and trading signal generation. Neuralk says predictions can be returned in seconds and that customer data is processed but not stored or used for training. It also offers public, reproducible benchmarks and enterprise compliance documentation. The company positions its approach as an alternative to traditional machine-learning workflows and AutoML, emphasizing rapid deployment and performance on structured data. Neuralk lists support from French Tech 2030, Microsoft for Startups, Station F, Inria, Scaleway, and Nvidia Inception Program.