"Lecture Scribe team on Vultr: Agent Rush Hackathon Hackathon"
Team Idea
I'm building Lecture Scribe, a voice agent for university classrooms. It records lectures and transcribes with an open ASR model, then fixes what ASR gets wrong: mumbled phrases, words lost to noise, garbled technical terms. The lecturer's slides and notes are the reference. Every repair is marked, nothing silently rewritten. Output is a clean transcript, lecture notes, and a list of places where speech contradicts the slides. The teacher approves everything before students see it. The agent runs sandboxed on Vultr, reads course material, writes only to a teacher-review queue. This builds on a hands-free voice agent I use daily to read course material aloud while I drive. Classes have 60+ students.
Team Members
.png&w=256&q=75)
KIRAN V K
kvk-code
Assistant Professor
Submission
Team Leader hasn't made a submission yet