Velo 3.0 - AI video infrastructure to explain, train, and sell faster.

by
Velo 3.0 is here, our third and biggest launch yet. Start with a screen recording or a prompt. Describe the video you need, and Velo writes the script, narrates it in your own voice, and builds the finished cut. It stays grounded in your company knowledge, from your docs and tools you connect through connectors and MCP. Edit by typing changes. Localize the finished video into 25+ languages in one click. Recording or prompt, Velo turns either into a polished video you can ship anywhere.

Add a comment

Replies

Best

the "grounds itself in your company knowledge" part is the interesting bet here, most video tools just take what you type or record at face value. question about that: once a video's generated from your docs/connectors, what happens when the source doc changes afterward, does the video get flagged as possibly stale, or is it a snapshot that's accurate as of generation time and then just quietly drifts from the source over time?

 Today, it’s a snapshot accurate at generation time. The sources stay attached for traceability. We’re building source-aware freshness so when a connected doc changes, Velo can flag the affected video and update only the relevant sections.

 Right now, it's more of a snapshot at generation time, but regenerating pulls in the latest context from all your connected sources

 makes sense, and good to see Sourav mention the flagging/partial-update piece is already on the roadmap. that's honestly the harder problem to get right, most tools would just leave it as a silent snapshot forever

Velo 3.0 feels less like a video editor and more like a teammate.


Give it a prompt, let it pull context from your knowledge sources, and it’ll create the video for you.


Need changes? Just tell it what to fix.


Hope you’ll give it a spin and tell us your thoughts!

Really looking forward to all the feedback on Velo 3.0

Loved Velo specially the part where i could natively change things thru natural language chat box. Kudos to the team on launch.

 Thank you so much

 Glad you liked it

The magic isn’t that Velo 3.0 makes videos.

It’s that it knows where to look before it makes them.

Grounding the agent in your company’s knowledge makes all the difference.


Can’t wait for everyone to try it!

 Lessgo

 Excited for the feedback from users!!!!

How do you handle varying audio quality in the input recordings, and is there a way to adjust or fine-tune the AI's editing decisions?

 You could do your own voice clone and edit every single word the AI generates in the edit script. On the prompt side of things you can just direct it however you want it to be

 Yep, you can edit every word yourself or let AI Rewrite do it in different personality styles

This is super cool! My only question is how accurate the voice cloning in and what the use cases are. Would I prompt Velo and they'd be able to clone my voice regardless of the context? Other than that I think this is gonna be super important in any workflow. Nice job!

you’ll have to give us a voice sample and you prompt the rest to your best output

Velo is something my team has been using since couple of weeks now and it has been amazing the speed at which we are shipping the product explainer videos to our clients.

 Appreciate it a lot

 Thanks you so much!

Velo has been a great product that helps me make product videos quickly

 Thank you so much!

 Thanks a lot!

Congrats!I’m curious whether the product is more focused on internal enablement videos, customer-facing demos, or turning existing knowledge into reusable video assets.

 thank you, today it's all three, honestly. We see teams using Velo for customer demos, internal training, onboarding, support, and product updates. The common thread is turning company knowledge into videos people can actually use

would love a way to auto-generate captions in different languages for the polished videos, that would save me a ton of time when sharing tutorials with international teammates.

 Added to feature roster. We can add this super easy

 We don't support it today, but I can definitely see how it'd save a lot of time