SBIR: Art of Novel Signals: Predicting and Forecasting with High Confidence

The contract solicitation titled SBIR Art of Novel Signals: Predicting and Forecasting with High Confidence seeks to advance predictive modeling through Temporal Knowledge Graph Forecasting and In-Context Learning using novel, multilingual, and multimodal data sources. It is issued under the Small Business Innovation Research program with a total small business set-aside, meaning only eligible small businesses with fewer than 500 employees may apply. The effort is sponsored by the Department of Defense through the Small Business Projects Office and is focused on developing high-confidence forecasting capabilities that can extract insights from complex, cross-domain signals, integrating temporal relationships, language diversity, and multiple data modalities. The solicitation was posted on July 22, 2026, with a firm response deadline of August 19, 2026, and the place of performance and point of contact details have not been specified. Interested parties must submit proposals through the official DARPA opportunity portal to be considered for funding under this SBIR mandate.