Ryan Gilbert

Ragie - Your Audio and Video, Now Fully Searchable.

Transcribe, index, & retrieve answers from spoken and visual content—with precise timestamps and streaming playback. Search recordings to find the moment something was said or shown. Base Chat now supports A/V, so users can ask and instantly watch answers.

Add a comment

Replies

Best
Bob Remeika

Hi Product Hunt! 😸

Today marks a huge milestone for us. Ragie now natively supports Audio and Video RAG!

Whether it’s a sales call, product demo, podcast, or company all-hands—you can now search and replay the exact context that matters.

With Ragie, you can:

🎙️ Upload an audio or video file

🔍 Ask a question about it

⏪ Stream the answer from the precise moment it was said

This is a huge unlock for developers building AI products with rich, immersive content. No need to stitch tools together or settle for transcripts alone—it works out of the box with the full Ragie pipeline.

And yes, Base Chat now supports audio and video too. You can chat with your knowledge base and jump straight to the relevant moments in recordings. No manual tagging or timestamping required.

We’re beyond excited to see what you’ll build with audio and video RAG. Let us know what you're cooking up! 🔥

– Bob

Supa Liu

Ragie sounds incredibly efficient for working with audio and video, allowing me to quickly find specific moments through transcription, indexing, and even asking questions.

Bob Remeika

@supa_l it is! Check it out and let me know what you think

Hossein Yazdi

I already knew Ragie since last year. It's really one of the most complete and advanced AI dev tools I've come across. Great job, guys. Upvoted you!

Matt Kauffman

@hosseinyazdi thank you! Appreciate it :) Reach out if you have any RAG use cases in your apps. We're always happy to help work out the best way to design AI features even if Ragie doesn't end being the best fit.

Serg

Is there a way to run it locally?

Matt Kauffman

@sergproduct the a/v pipeline and retrieval system can't be run locally, but the app Bob is using in the demo video to test retrievals is open source and can be run run locally. You can find it here: https://github.com/ragieai/promptie.

Ohans Emmanuel (remention.ai)

Brilliant, congrats on the launch!

Matt Kauffman

@ohansemmanuel thank you :)

Ryan Gilbert

Congrats on the launch!

What was the biggest technical challenge in getting this shipped?

Matt Kauffman

@ryangilbert there were quite a few! I'll let some of the other team members speak to specific areas they worked on. My focus was around processing media files to prepare them for semantic extraction and streaming and also securely streaming media from the segment where a retrieval match occurs. In other words if the best section of a video to answer a query is 2:30 in, we provide a way to stream starting at that timestamp, while ensuring only authenticated requests can access the content.

qilong

Congrats on this major update! Native support for audio and video RAG is a real game-changer, especially for teams dealing with lots of recorded content. I’m curious—how does Ragie handle different accents or background noise in audio files? Also, are there plans to support real-time audio/video streams in the future, or integrations with popular meeting platforms like Zoom or Teams? Super excited to see how developers use this!

Bob Remeika

@qilong thanks for checking us out. We use models that filter these accents and background noises out. We definitely plan to support realtime a/v and platforms like Zoom and Teams in the very near future.

Bennett Cohen

any plans on making this fully open source? looks sweet

Bob Remeika

@bennett_cohen1 we probably wont make the pipeline fully open source in the near term but we do have a lot of open source tools and projects including our multi-tenant chat bot (Base Chat)

https://github.com/ragieai/basechat/

orliesaurus

Amazing release, I wanna add this to my doc search so I can index mp3s

Bob Remeika

@orliesaurus it's great for that! I hope you like it