Fast-response virtual teams (FRVTs) have been developed as a response to emergent challenges faced by organizations that need to be addressed urgently. Even though FRVTs offer enormous potential in terms of their benefits, their success is not guaranteed. When used, the need for high performing FRVTs has become critical for organizational success. However, there is a lack of detailed understanding of how sensemaking can potentially influence FRVT performance. Drawing on social exchange theory, we identify swift trust as a potential antecedent of sensemaking. In this paper, we report the results of a study that examined the effects of swift trust on sensemaking and the effects of sensemaking on team performance in FRVTs. The study included 20 FRVTs and 80 team participants. Analysis of data shows that FRVTs’ swift trust is positively correlated with all three dimensions of sensemaking and only the linguistic and conative development aspects of sensemaking affects FRVT performance.

Homomorphic encryption (HE) is the ultimate tool for performing secure computations even in untrusted environments. Application of HE for deep learning (DL) inference is an active area of research, given the fact that DL models are often deployed in untrusted environments (e.g., third-party servers) yet inferring on private data. However, existing HE libraries [somewhat (SWHE), leveled (LHE) or fully homomorphic (FHE)] suffer from extensive computational and memory overhead. Few performance optimized high-speed homomorphic libraries are either suffering from certain approximation issues leading to decryption errors or proven to be insecure according to recent published attacks. In this article, we propose architectural tricks to achieve performance speedup for encrypted DL inference developed with exact HE schemes without any approximation or decryption error in homomorphic computations. The main idea is to apply quantization and suitable data packing in the form of bitslicing to reduce the costly noise handling operation, Bootstrapping while achieving a functionally correct and highly parallel DL pipeline with a moderate memory footprint. Experimental evaluation on the MNIST dataset shows a significant ( $37\times$ ) speedup over the nonbitsliced versions of the same architecture. Low memory bandwidths (700 MB) of our design pipelines further highlight their promise toward scaling over larger gamut of Edge-AI analytics use cases.