Recording
Voice, video, and screen captures pinned to the exact spot in the work.
Stop pushing your users to another tool every time they need to show, not tell.
Free tier. No credit card. First comment in 5 minutes.
The real shift this quarter is that agents ship alongside the team, not after them.
Right, and that’s exactly where recording comes in for async review.
This sounds a bit flat — try a warmer take here.
This sounds a bit flat — try a warmer take here.
Jump-cut here — the file dialog stalls for a sec.
Dead air on this frame — trim 0:12–0:15?
Native recordings running inside products at
What it is
Voice, video, and screen recordings pinned to the exact spot in the work. Recordings are Velt objects: commentable, notifiable, audit-trailed. The Velt Recorder is a set of components: Recorder Notes pin recordings to specific locations on the screen, the Recorder Tool starts one, the Control Panel manages it, the Player plays it back by ID, and the Video Editor edits it in place. Recordings transcribe automatically, so what was said becomes text that humans and agents can read.
“The variance comes from the vendor rate change, see row 18.”
Confirmed: the rate change in row 18 explains 92% of the variance. Suggested correction attached.
one recording, both actors, the approval visible on the work
How it works
Wrap your app, mount the recorder components, and voice, video, and screen recording work out of the box. Recordings pin to the work, transcribe automatically, and play back anywhere through the Player.
npm install @veltdev/react<VeltProvider
apiKey={VELT_API_KEY}>
<YourApp />
</VeltProvider><VeltRecorderNotes />
<VeltRecorderTool type="all" />
<VeltRecorderControlPanel />
<VeltRecorderPlayer />Mount the Recorder Tool where users work and set its type to audio, video, screen, or all. Recordings pin to on-screen locations through Recorder Notes and play back anywhere through the Player. Transcription runs automatically and can be turned off. The SDK fires an event at every step of the recording lifecycle, and recordings are queryable through the SDK and the Get Recordings REST API. The video editor opens on a finished take without leaving the page.
// type='audio' | 'video' | 'screen' | 'all' · transcription on by default
Teams that build it budget a quarter for the recorder alone; the editor and transcription are each their own projects. The 3 steps above replace all three.
Skip the steps. Have your agent set it up.
npx -y @velt-js/mcp-installer
claude mcp add velt-installer -- npx -y @velt-js/mcp-installer
npx -y @velt-js/mcp-installer
npx -y @velt-js/mcp-installer
npx -y @velt-js/mcp-installer
Self-hosted media
APIs and data
Launch recordings this week.
No credit card. Works with React, Next.js, Vue, Angular, and HTML.
Capabilities
Each card is the live SDK. Toggle to Code for the exact snippet that renders it.
This number is faster said than typed: the variance is the vendor rate change, see row 18.
Pin this to C18 so finance sees it on the cell.
React
Next.js
Angular
Vue
HTML
Google Cloud Storage
Azure Blob<VeltRecorderTool type="audio" />One tap starts an audio note; the clip pins to the cell or clause it is about. When the explanation is faster spoken than typed, it stays where the number is.
Quick walkthrough of the pricing slide.
<VeltRecorderTool type="video" />Camera recordings captured and saved inside your product, played back in the thread. The reviewer’s face and tone travel with the deck instead of a pasted link.
Jump-cut here — the file dialog stalls.
<VeltRecorderTool type="screen" />Tab or full-screen recording with mic narration, started from one button in your UI. A reviewer walks through the filing the way they would in person.
“The variance comes from the vendor rate change, see row 18.”
<VeltRecorderNotes />Recorder Notes anchor every recording to a position on the screen: the cell, the field, the frame. Show-not-tell feedback lands on the work, not in a separate library.
“…see row 18 for the vendor rate change.”Key point: the variance is driven by the vendor rate change (row 18).
// transcription on by default
<VeltRecorderPlayer />Recordings transcribe automatically, with subtitles and an AI summary on the player; one prop opts out of the LLM. A three-minute walkthrough becomes skimmable text.
Dead air on this frame — trim 0:12–0:15?
// the editor opens on the finished take
<VeltRecorderTool videoEditor />No more re-recording a take to fix the first ten seconds: trim, split, zoom, and delete segments right where the recording was made, no export, no third-party tool.
Recorded the walkthrough: see the filing steps here.
// a recording attaches like any Velt object
<VeltComments />A recording attaches to a comment thread like any other Velt object: reply, resolve, deep-link. The walkthrough and the decision it triggered live in one place.
velt.on()velt.on("recordingDone", handler);The SDK fires an event at every step: started, paused, done, transcription ready, edit complete. Drive your own UI, analytics, or audit pipeline from a recording’s whole life.
Only minimal identifiers (recordingId, userId, timestamps), so the media never leaves your storage.
// recorder data provider routes media to your storageThe recorder data provider keeps recorded files, identity, transcription, and attachment URLs on your storage; Velt stores minimal identifiers. The compliance answer when users record regulated work.
Our reviewers record a 40-second walkthrough instead of typing a paragraph nobody reads. The feedback lands on the exact cell, and the transcription makes it searchable later.
Head of product · FP&A platform
Little big details
Shipped functionality only. Every item traces to the recorder docs; draft items render after engineering sign-off. This is the part of an in-house build that never ends.
Make it yours
Prebuilt recorder components for the fast path, wireframes for fully custom UIs, and configuration, events, data APIs, and a recorder data provider underneath.
Every recorder surface is wireframe-customizable: Recorder Tool, Control Panel, Media Source Settings, Player, Player Expanded, Recording Preview Steps Dialog, Subtitles, Transcription, and Video Editor. Template variables and dark mode apply.
“the variance is the vendor rate change”
<VeltRecorderToolWireframe>
// your recorder markup
</VeltRecorderToolWireframe>Recording type per tool instance, floating vs thread mode, max length, quality and encoding controls, transcription and summary toggles, editor auto-open and retake, the full event stream, data APIs, the Get Recordings REST API, and the recorder data provider for storage routing.
<VeltRecorderTool
type="all" recordingTranscription={false}
/>We routed recorded files to our own S3 with the recorder data provider and kept the prebuilt player. Customizing the editor and subtitles to match our brand was a config change, not a project.
Staff engineer · healthcare SaaS
In production
Tabbed by vertical, with verified customer screenshots.
This Digital Sales Room gives you everything you need to move forward — from proposal to plan.
@Fin We need to make a working demo for Rene and his team
A strategist records a 60-second walkthrough pinned to slide 4; the client watches and replies in the thread. The approval happens on the deck, with no call scheduled.For sales enablement
finance / receivables
Can we confirm the PO before approving this one? @Sarah
1 ReplyThe variance explanation is a voice note on the Q3 cell, transcribed to text. Months later, the auditor reads what was said, when, and by whom.For fintech and FP&A
Cotton Supplier Ltd.
AgriCotton Farms
Village Kheda, Gujarat, India
SpinCo Yarns
18 Narol Rd, Gujarat, India
DyeWorks International
Calle de Mayo, Mexico
Can you attach the GOTS cert for this lot before we approve? @Sean
1 ReplyA field tech records the damaged shipment from a phone, pinned to the order record. The claim decision happens on evidence, not on a description of evidence.For operations
Privacy Risk Review
Wren discovered a feature request in Jira involving users' health data and triggered a review after analyzing the PRD in Confluence against company policy and US regulation.
JStarted by Juan MendezThis feature processes PHI, so HIPAA applies. I recommend a PIA and a vendor BAA review before any data flows. @Juan
The officer records the attestation rationale as a voice note pinned to the control, transcribed to text. Months later, the examiner reads why the disclosure was approved, when, and by whom.For compliance
Superflow
1 · Subscription & Fees
Customer subscribes to the Service on a monthly basis and shall pay the then-current fees in advance of each billing period. Fees are non-refundable except as expressly set out in this Agreement.
2 · Term & Auto-Renewal
The subscription renews automatically for successive one-month terms unless either party gives at least thirty (30) days' written notice of non-renewal before the end of the then-current term.
3 · Limitation of Liability
Except for breaches of confidentiality, each party's total aggregate liability under this Agreement shall not exceed the fees paid by Customer in the one (1) month preceding the event giving rise to the claim.
4 · Governing Law
This Agreement is governed by the laws of the State of Delaware, without regard to its conflict-of-laws provisions, and the parties consent to the exclusive jurisdiction of its courts.
Counsel records why the clause was accepted as a 60-second walkthrough pinned to the matter. The next reviewer reads the rationale instead of guessing at the redline history.For legal
Make the hero headline punchier and add a primary CTA.
Done — rewrote the headline, tightened the subtext, and added a primary CTA in hero.tsx.
Add a logo strip with social proof under the CTA.
Added a five-logo trust strip below the CTA and wired it into hero.tsx.
Ship your app in a weekend
Get startedUse our brand orange on the CTA and cut the headline to one line. @Sean
A reviewer records why a generated draft was rejected; transcription turns the rationale into text the next reviewer and your agents can read.For AI-native SaaS
See it running in products like yours.
30 minutes, with an engineer, not a sales deck.
Related primitives
The walkthrough lands in a thread, goes live in a huddle, or pings the right reviewer.
A recording can live inside any thread; the thread holds the decision.
Replied on the pinned recording
voice note · 0:42a recording lives inside a thread the thread holds the decision
When async recording is not enough, the conversation goes live in the same document.
async recording not enough go live in the same document
The pipeline that tells a reviewer a recording landed.
Hope shared a recording
a recording lands the reviewer is told
Recorded files, user identity, transcription, and attachment URLs can live on your infrastructure via the recorder data provider, with only minimal identifiers on Velt servers. Recording runs on the same isolation guarantees as the rest of the SDK. See self-hosting and governance.
Book DemoProof
Show-not-tell feedback, captured where the work is.
0 tools
Our users used to leave for a screen recorder and paste a link back. Now they record in place, pinned to the cell, and never leave the product to show what they mean.
1 afternoon
We mounted the recorder components and had voice, video, and screen recording working that afternoon. The transcription and editor were already there.
40s
A 40-second walkthrough replaced a thread of typed back-and-forth. The transcription makes it searchable months later when the auditor asks.
FAQ
Install @veltdev/react, wrap your app in VeltProvider with your API key, and add the recorder components: VeltRecorderNotes for pinned recordings, VeltRecorderTool to start one, VeltRecorderControlPanel and VeltRecorderPlayer for control and playback. Voice, video, and screen recording work out of the box; the quickstart walks through it.
Yes, all three from one Recorder Tool: set type to audio, video, screen, or all. Screen recordings support mic narration, and Picture-in-Picture in Chrome when the camera is on.
Yes. AI transcription is on by default, with subtitles and an AI-generated summary on the player. One prop turns it off, and disabled means disabled: the recording is never sent to an LLM.
Yes. The built-in video editor opens on the finished take: trim, split, zoom, and delete segments on a timeline with frame previews, for video and screen recordings. It can open automatically when recording stops, and users can retake from inside it.
By default, Velt stores them. With the recorder data provider, recorded files, user identity, transcription, and attachment URLs stay on your storage (any provider: S3, Google Cloud Storage, Azure Blob, or your own servers) while Velt keeps only minimal identifiers. See /self-hosting.
Yes. Voice, video, and screen recordings attach to comment threads, and a recording links to a comment annotation under the hood, so feedback about a recording stays with the recording. See /comments.
Yes. Velt is optimized for mobile web, works inside WebViews in native apps, and pure native apps can integrate through the REST APIs with your own native UI.
Velt is priced on usage, not seats: you pay for documents with review activity in a month, and there is a free tier for development and early production. Recording is part of the SDK, not a separately priced add-on.
Free tier. No credit card. First comment in 5 minutes.
30 minutes, with an engineer, not a sales deck.
@Jordan Auto-renewal needs 60 days' notice to match our policy, not 30.
@Maya Liability cap is one month of fees — Legal wants 12 months before we sign.
@Maya Governing law should be New York to match the MSA.