Visa's newly granted patent describes a way for two devices to complete a payment without an internet connection, using cryptography to verify each transaction. According to the patent, a device determines whether a payment "amount is less than an offline amount stored in the secure element," updates that stored balance, and returns a cryptographically signed "interaction response message" to the other device. Dallas-based Mustafa Ozdayi is among 12 inventors named on the patent. [Composite image: Sources, U.S. Patent No. 12,647,284-B2, Fig. 7; DI Studio]
Dallas-based Verily Health Inc.'s newly granted patent describes a virtual-reality headset designed to automate the pupil-response exam an eye care professional typically performs by hand with a penlight, using an infrared eye-tracking subsystem and light-sealed compartments over each eye to measure how the pupils react to controlled light stimuli. According to the patent, the manual version of the test produces results that "vary by practitioner," and Verily's system is designed to improve the "sensitivity, consistency, and ease" of conducting it, as well as to flag asymmetric pupil sizes — anisocoria — alongside afferent pupil defects. Fig. 1 shows the system in use: a user wears the headset while a clinician runs the test from a tablet. [Composite image: Sources, U.S. Patent No. 12611100, Fig. 1; DI Studio]
Charles Herder (left) and Tina Srivastava (right), co-founders of California-based privacy company Badge Inc., are named on U.S. Patent No. 12,574,371 for a system that authenticates users by splitting biometric data into “shards” stored across multiple servers, so no single location holds the complete identifier. Both founders have innovation ties to Dallas. Pictured with them is Dan Kaufman (center), a cybersecurity veteran known as “DARPA Dan,” who joined Badge in 2024 as SVP of AI and cybersecurity. The company, which emerged from stealth in 2024, was named to TIME’s Best Inventions of the Year and Fast Company’s Innovation Team of the Year in 2025. [Photo: Badge Inc.]
The Board of Regents of the University of Texas System, Stanford University, and the U.S. Department of Veterans Affairs have received a newly granted patent for a method that uses machine learning to help identify which people with depression are likely to respond to antidepressant treatment. The patent lists Dallas-based Madhukar Trivedi among its inventors and is assigned to the three institutions. [Composite illustration: Sources, USPTO Patent #12490933; DI Studio]