
3rd Place Grand Award, Software Design · Regeneron ISEF 2026
DeepScope
A deepfake detector built on the involuntary signals of a real face, like micro-tremors and blinking, instead of the artifacts one generator leaves behind. Generators aren’t trained to fake biology, so a detector that reads it should keep working on generators it has never seen. It started as my science fair project, and since ISEF I’ve rebuilt it from scratch to put that idea to a harder test.
- 1st Place, Twin Cities Regional Science Fair 2026
- ProEquip Award for Safety in AI, 2026












