Launched this week
DocQuest
Document insight, AI tutor, pdf, audio & video to podcasts
9 followers
Document insight, AI tutor, pdf, audio & video to podcasts
9 followers
DocQuest transforms PDFs, audio & video into intelligent podcasts. Unlike fragmented tools, we unify all formats in one platform. ✨ Multi-doc analysis (10 files at once) ✨ PDF, audio, video → Podcast ✨ 24/7 AI Agents for automation ✨ Adaptive AI Tutor for learning
The audio/video to podcast pipeline is a really compelling use case. I'm building in a similar space (turning creative input into different media formats with AI) and the multi-format ingestion is the part that usually trips people up. How do you handle long-form video content — does it chunk it into segments or try to process everything at once? And does the podcast output include generated narration or is it more of a restructured summary?
@mattias_s Great question! For long-form video, we extract the audio track first, then pass it
to our model for transcription. we splits large audio into 10 mins and chunks with 5s overlap, then transcribes each at full fidelity, and concatenates to avoid lossy compression
On the output side it's fully generated narration, not a summary.
The multi-format ingestion piece you mentioned is real we support direct file upload (audio/video), remote URLs, YouTube links and text/document collections. All feeding into the same generation pipeline.
Curious what format outputs you're targeting in your project?
@maximilien_kpizingui Thanks for the detailed breakdown! The 10-min chunks with 5s overlap is smart — avoids the boundary artifacts you'd get from hard cuts. Good to know the output is full narration rather than a summary, that's much more useful.
For my project (MusicOrb) the output is MP4 music videos — we chain Claude for lyrics/storyboard, Suno for audio generation, Gemini for scene images, and Remotion for video compositing. Different domain but similar chunking challenges when syncing lyrics to audio timestamps. Would love to see DocQuest handle music/audio analysis too someday!
Turning a dense PDF into something you can listen to while commuting is a genuinely useful idea.
@utkarsh_singh64 ,yes indeed thank you for your upvote