LectureVoice turns lecture slides into a narrated video, in the lecturer’s own voice.
I built it for a professor at University of Sri Jayewardenepura (a state uni in Sri Lanka) who was spending whole evenings recording lectures. Now he uploads his slides and a short voice sample, and gets back a finished video with subtitles. He can edit the script for each slide and redo any single slide before it’s rendered.
I built it end to end: Next.js frontend, Python/FastAPI backend on AWS, PostgreSQL, and GPU jobs on Modal that train a small voice model for each user. Voices are never shared, and every clip has a hidden AI watermark.
Live since June 2026.
0 likes · 1 view · Oct 26, 2026
LectureVoice turns lecture slides into a narrated video, in the lecturer’s own voice. I built it for a professor at University of Sri Jayewardenepura (a stat...
