# Recording, Meetings & Podcasts

CoreReflex isn't only for AI-generated film — it's also a full capture booth, a text-based podcast editor, and a real-time voice agent. You can record your screen, your camera, your microphone, and the audio playing on your computer, all from inside the studio, with nothing to install. Whatever you capture lands straight on the timeline, where the tools that matter most for spoken-word content are waiting: word-accurate captions with one-click styles and translation, a transcript you can *edit like a document* (delete the words, and the video cuts itself), one-knob mastering, a multitrack Audio Studio with broadcast loudness targets, and an Auto-Clips studio that turns a long episode into ranked vertical shorts. And when you want a recording to *think back* at you, the real-time voice agent can answer callers, qualify leads, handle grounded support questions, and book appointments out loud.

**On this page**

- [Where to record](#where-to-record)
- [The in-browser recorder](#the-in-browser-recorder)
- [Podcasts: studio sound in one click](#podcasts-studio-sound-in-one-click)
- [Auto-captions: styles, animations, translation, and subtitle files](#auto-captions-styles-animations-translation-and-subtitle-files)
- [Edit the episode by editing the transcript](#edit-the-episode-by-editing-the-transcript)
- [Auto-Clips: turn a long recording into vertical shorts](#auto-clips-turn-a-long-recording-into-vertical-shorts)
- [Master the mix in the Audio Studio](#master-the-mix-in-the-audio-studio)
- [Turn a recording into a meeting brief](#turn-a-recording-into-a-meeting-brief)
- [The real-time voice agent](#the-real-time-voice-agent)
- [The Chrome extension: quick tab capture](#the-chrome-extension-quick-tab-capture)
- [Putting it together](#putting-it-together)

When you're done capturing, the [Editing & Mastering](editing-and-mastering.md) guide takes you the rest of the way to a finished cut, and [Meetings & AI Briefs](meetings-and-briefs.md) turns a recorded conversation into a structured summary.

---

## Where to record

The recorder lives in the **studio editor** at [editor.corereflex.com](https://editor.corereflex.com). Open a composition (or start a new one from the studio hub), and find the **Record** panel in the inspector on the right. From the dashboard at [corereflex.com](https://corereflex.com), the **Record screen + cam** and **Record a podcast** cards both drop you into the editor with the recorder ready.

Everything you capture uploads to your own workspace storage — the same private fleet your generated clips land on — and appears under your imports in the editor. It never passes through a third-party recording service.

---

## The in-browser recorder

The recorder uses your browser's native capture, so it works without any plugin or desktop app. You pick a source, choose which audio to include, and hit record. When you stop, the clip is saved and placed on the timeline automatically.

### Choose what to capture

The Record panel gives you two sources:

| Source | What it captures |
|--------|------------------|
| **Screen** | A screen, a window, or a single browser tab — your browser asks which one when you start. Ideal for screencasts, demos, and meetings. |
| **Camera** | Your webcam. Ideal for talking-head intros, podcasts, and pieces to camera. |

Two audio toggles sit beneath the source picker:

- **Microphone** — records your voice from your mic. On by default.
- **System audio** — records the sound your computer is playing (music, a call, a video you're sharing). This option only appears for **Screen** capture, since there's no system audio to mix into a plain camera shot. On by default for screen recordings.

If you turn on both Microphone and System audio, CoreReflex mixes them into a single clean audio track for you — no separate stems to wrangle.

### Record and place

1. In the **Record** panel, pick **Screen** or **Camera**.
2. Tick **Microphone** and, for screen capture, **System audio** as needed.
3. Click **Start recording**. Your browser will prompt you to allow camera or screen access and, for screen capture, to choose the screen, window, or tab to share. Grant the permission to begin.
4. The button turns into a live **Stop** timer so you can see how long you've been rolling. Recording also stops cleanly if you click your browser's own "Stop sharing" control.
5. Click **Stop**. CoreReflex finishes the recording, uploads it, and drops it onto the timeline as a selected clip, ready to edit.

If your browser blocks access, you'll see "Permission denied — allow screen/camera access to record." Re-grant the permission in your browser's site settings and try again.

### What you can do with the clip

Because the recording is just another clip on the timeline, every editing tool applies to it:

- **Trim and arrange** it on a multi-track timeline alongside generated shots, voiceover, and music.
- **Caption** it — and then **edit it through the transcript**: delete words, cut fillers, trim pauses (see below).
- **Master the audio** with the one-knob FX rack, or take the whole mix into the [Audio Studio](#master-the-mix-in-the-audio-studio).
- **Color-grade** it, add **transitions**, and **render** it out.
- **Brief** it — turn a recorded meeting into an AI summary via [Meetings & AI Briefs](meetings-and-briefs.md).
- **Clip** it — turn a long episode into ranked vertical shorts with [Auto-Clips](#auto-clips-turn-a-long-recording-into-vertical-shorts), or reframe a finished cut with the [Shorts Studio](editing-and-mastering.md#shorts-studio).

For the full timeline walkthrough, see [Editing & Mastering](editing-and-mastering.md).

---

## Podcasts: studio sound in one click

A podcast is just a recording — usually camera plus mic, or a screen share for a remote conversation — followed by a little audio polish. CoreReflex handles the polish for you.

### Record the podcast

Use the recorder above with the **Camera** source and **Microphone** on. For a remote co-host, share the call as a **Screen** recording with **System audio** enabled so both voices are captured, then mix down to one track automatically.

### "Podcast voice" mastering

Master audio FX is a one-knob-per-effect mastering rack for your whole mix, found in the editor's master audio controls. Instead of learning what a de-esser or limiter does, you pick a **preset** and it dials every knob for you. The **Podcast voice** preset applies a chain tuned for spoken-word recordings:

| Effect | What it does |
|--------|--------------|
| **Compress** | Evens out loud and quiet passages so the voice stays consistent and punchy. |
| **De-ess** | Tames harsh "sss" sibilance. |
| **Denoise** | Reduces background hiss and hum. |
| **Limit** | Holds peaks under a safe ceiling so nothing distorts. |
| **Maximize** | Pushes overall loudness up to a broadcast-ready level. |

Apply **Podcast voice** and you're done — but every knob stays adjustable, so you can nudge any effect to taste. Other presets are available too, including **Dialogue**, **Clean voice**, **Music master**, and **Punchy**.

These effects are real audio processing, not a simple volume curve, and they're **baked into your export** when you render — so the master you ship sounds exactly like the one you previewed. The full effect list and the per-knob controls are covered in [Editing & Mastering](editing-and-mastering.md).

Want more than one knob per effect — a multitrack mix, music that automatically dips under your voice, and a precise loudness spec for each platform? That's the [Audio Studio](#master-the-mix-in-the-audio-studio), below.

---

## Auto-captions: styles, animations, translation, and subtitle files

Any recording — or any clip on your timeline — can be captioned automatically. CoreReflex listens to the audio and transcribes it with **word-level timestamps**, then adds the words as their own animated caption layer above the clip. Because every word carries its own timing, each word can light up exactly in time with the audio — and, just as importantly, the transcript becomes an *editing surface* (next section).

### Generate captions

1. Select the clip on the timeline.
2. In the inspector, open the **Captions** section.
3. Click **Caption video** (or **Caption audio**). Transcription runs in the background; a task indicator shows progress.
4. The finished captions appear as a styled, word-highlighted layer on the timeline, directly above the clip.

A few things worth knowing:

- Captioning is available for **audio and video** clips.
- Transcription is **US English** by default and adds punctuation automatically.
- There's a length limit of roughly **13 minutes of audio** per clip. If a clip is over it, the button is replaced with "Audio is too long to caption" — split a long recording into shorter pieces and caption each one.
- If captioning is unavailable because the workspace isn't fully configured, you'll see a clear "Captions are disabled" message rather than a silent failure.
- You can **Regenerate captions** at any time if you re-record or edit the source.

### One-click caption styles

Select the caption layer and open its **Style** section in the inspector. Four presets restyle the whole layer in a single click — and a single undo:

| Preset | The look |
|--------|----------|
| **Karaoke** | Big bold words with a blue active-word highlight and a pop on each word — the classic social-video style. |
| **Bold pop** | One huge line at a time, gold highlight, heavy stroke — maximum punch for shorts. |
| **Minimal** | Clean, stroke-free medium text with a gentle rise on each word. |
| **Subtitle** | Smaller, restrained two-line captions — the broadcast-subtitle look. |

Presets only touch the *look* (size, colors, stroke, weight, lines, animation) — never your words or timing. After applying one, every individual control stays live: font (TikTok Sans by default), size, position, line count, page duration, fill and **highlight color** for the active word, stroke, and more, spread across the Layout / Typography / Fill / Stroke / Captions sections of the inspector.

### Word animations

Right under the style presets, pick how the **active word** enters:

- **None** — the highlight color changes with no motion.
- **Pop** — each word lands slightly oversized and settles, karaoke-style.
- **Rise** — each word slides up a few pixels into place.

The animation you see in the preview is exactly what renders in the export — it's computed the same way in both.

### Translate the captions

In the caption layer's **Transcript** section, a **Translate** row converts the whole caption track into another language:

1. Pick a target language — **Spanish, French, German, Portuguese, Italian, Japanese, Korean, Chinese, Hindi, or Arabic**.
2. Click **Translate**. The translated words are re-timed onto the same points in the video, so they stay in sync with the speech.
3. Not happy? **Undo restores the original language instantly** — the original captions are never destroyed.

### Download SRT / VTT subtitle files

At the bottom of the **Transcript** section, the **Export sidecar** row gives you the transcript as a standard subtitle file:

- **Download SRT** — the universal subtitle format (YouTube, most players and platforms).
- **Download VTT** — the web-native format (HTML5 video, many podcast hosts).

These are sidecar files: they don't change your project, they just let you carry your transcript anywhere — upload it alongside the video, hand it to an editor, or archive it.

---

## Edit the episode by editing the transcript

This is the heart of the podcast workflow: once a clip is captioned, you edit the *words*, and CoreReflex edits the *video*. Deleting a sentence from the transcript cuts that stretch out of the source clip, closes the gap, and keeps the remaining captions perfectly in sync — all in one undoable step.

Select the caption layer on the timeline and open the **Transcript** section in the inspector.

### Click a word to jump there

Every word in the transcript is clickable. Click one and the player **seeks straight to that moment** — no scrubbing the timeline hunting for the sentence you half-remember. Reviewing an hour-long take becomes reading, not watching.

### Delete words — the video cuts itself

1. Click the first word you want gone.
2. **Shift-click** the last word — the whole run lights up.
3. Click **Delete N words**.

CoreReflex removes that time range from the **source clip** (the pieces ripple closed with no gap), compresses the captions to match, and confirms exactly how much it cut — e.g. *"3.2s cut — the clip rippled closed and captions stayed in sync."* One **Undo** (Cmd/Ctrl+Z) brings it all back.

### Remove fillers and pauses in one click

The transcript tints likely **filler words** ("um," "uh," "erm," "hmm"…) in amber so you can see the clutter at a glance. Two chips beneath the transcript clean it all at once:

- **Remove fillers (N)** — cuts every detected filler word. This uses a conservative list of non-words only, so it won't eat words that carry meaning.
- **Remove pauses (N)** — trims every silence longer than about a second, with a little breathing room kept around the speech so cuts don't sound clipped.

Want finer control — different pause thresholds (0.6s to 3s) or the **Aggressive** filler list (*like, you know, basically, literally…*)? Select the **clip itself** (not the caption layer) and use the **Clean up audio** panel, which previews exactly what it will remove before you commit. Full walkthrough in [Editing & Mastering](editing-and-mastering.md#tighten-the-delivery-remove-filler-words-and-silences).

### Find the best moments with AI

Still in the Transcript section, click **Find highlights**. CoreReflex reads the transcript and returns its **best standalone moments**, each with a title, a score, a time range, and (on hover) the reason it was picked.

- **Click a highlight** to select its words in the transcript and jump the player there.
- **Keep only this** cuts everything *outside* the highlight — the whole composition ripples down to just that moment, in one undoable step. Perfect for pulling a teaser or a quote clip from a full episode.

For turning a long recording into *several* finished vertical shorts in one pass — cut, captioned, rendered, and even scheduled to socials — use [Auto-Clips](#auto-clips-turn-a-long-recording-into-vertical-shorts) instead.

### If transcript editing is grayed out

Caption layers made before transcript editing existed aren't linked to their source clip, so the delete buttons won't show — you'll see a note explaining this. Click **Regenerate captions** on the clip and the fresh caption layer will be fully editable.

---

## Auto-Clips: turn a long recording into vertical shorts

A podcast episode, a webinar, a livestream VOD — the long recording is the raw material; the shorts are what travel. **Auto-Clips** watches the whole video, finds the most engaging moments, ranks them, cuts each one to vertical, optionally adds styled captions and a voiceover, renders every short straight into your library, and can even schedule them to your social accounts on a cadence.

Open it at **[corereflex.com/auto-clips](https://corereflex.com/auto-clips)**, or from the **Auto-Clips** card on your dashboard.

### Step 1 — Pick a source and set the options

1. **Pick a video from your library** — your uploaded and generated videos appear as a thumbnail grid (sign in to see them) — **or paste a public video URL** (https).
2. Set the options:

| Option | What it does |
|--------|--------------|
| **Clips** | How many shorts to look for (up to 20; 4 by default). |
| **Min sec / Max sec** | The allowed length of each short (12–60 seconds by default). |
| **Aspect** | **9:16** (TikTok / Reels / Shorts), **1:1** (square feed), or **16:9**. |
| **Moments** | What counts as a highlight: **General** (hooks, insights, laughs, quotable moments) or **Gaming** (plays, reactions). |
| **Captions** | **None**, **Auto** (matches each moment), or a specific personality: **Gaming**, **GenZ**, **Funny**, **Dramatic**, **Minimal**, or a **Story** narrator (**News**, **Roast**, **Creepypasta**, **Dramatic**). |
| **Voiceover** | **On** adds a spoken narration matched to your caption style — and automatically turns the original audio down underneath it so the narration stays clear. |
| **Story mode** | **On** writes **one narrative arc across all your selected clips** (part 1, part 2, …) instead of independent captions per clip. Needs at least two selected highlights. |

3. Click **Detect highlights**.

Long videos are automatically prepared for analysis first — you'll see the progress as it works. If preparation isn't available on your deployment, short clips still analyze directly; you may be asked to enter the source's width, height, and duration before generating.

### Step 2 — Review the ranked highlights

Each detected highlight shows its **title**, **time range**, a **type** tag (hook, insight, laugh…), and a **score**. Everything is ticked by default — untick any moment you don't want, then click **Generate selected shorts**.

### Step 3 — Your shorts render into your library

Watch each card go **Queued → Rendering → Ready**. Finished shorts play right on the page with an **Open / download** link — and every one also lands in your **library**, so you can open any short in the editor and keep polishing (restyle captions, trim, regrade) like any other project.

### Step 4 — Distribute to socials

Once at least one short is ready, the **Distribute** panel appears:

1. Tick the **connected social accounts** you want to publish to. (No accounts yet? Connect TikTok / Reels / Shorts from your dashboard first.)
2. Optionally set a **start time** and an **every N hours** cadence — the shorts are staggered on that schedule. Leave start empty and hours at 0 to publish immediately.
3. Click **Schedule finished shorts** and CoreReflex confirms how many were scheduled.

### Auto-Clips or the transcript editor?

- **Auto-Clips** is the volume play: one long video in, a batch of ranked, captioned, rendered verticals out — no timeline required.
- **Find highlights** in the [transcript panel](#find-the-best-moments-with-ai) is the surgical play: pick one moment, "Keep only this," and hand-finish it in the editor.

They complement each other — many creators run Auto-Clips for the batch, then use the editor for the one clip that deserves extra love.

---

## Master the mix in the Audio Studio

For a quick single-track polish, the [Podcast voice preset](#podcast-voice-mastering) is all you need. When the episode has *layers* — narration, music bed, sound effects, a recorded co-host — switch to the **Audio Studio** for a proper multitrack mix.

In the editor's top bar, flip the mode toggle from **Video** to **Audio**, then open (or create) an audio session. Inside a session you get multitrack lanes, a mixer with level meters, record-into-track, and shortcuts to generate music or voiceover straight into the session. Two tools matter most for podcasts:

### Auto-duck: music that gets out of the way

Click **Auto-duck** in the session toolbar and CoreReflex writes a volume dip onto **every music track** wherever **voice** is playing — the music drops about 10 dB under the speech with a smooth quarter-second ramp in and out, and recovers between voice regions. You'll see a confirmation like *"Ducked 1 music track under 4 voice regions."*

- The dips land as ordinary **volume automation keyframes**, so you can hand-tune any of them afterward — move a ramp, soften a dip, delete one entirely.
- It's one **Undo** away from gone.
- Tracks are told apart by their clip names — anything CoreReflex generated (voiceover, soundtrack, mic recordings) is already named so this just works. For imported files, keep words like *music*, *song*, or *beat* in music filenames and *voice*, *podcast*, or *dialog* in speech filenames so they're classified correctly. If the session has no recognizable voice + music pair, the button tells you plainly instead of guessing.

### Loudness targets: hit the spec every platform expects

Next to the **Bounce** button, the loudness selector normalizes your final mix to a platform standard, measured the same way streaming platforms and broadcasters measure it (integrated LUFS):

| Target | Use it for |
|--------|-----------|
| **−14 LUFS · streaming** | YouTube, Spotify, and general streaming. |
| **−16 LUFS · social / podcast** | Podcast feeds and social platforms. |
| **−23 LUFS · broadcast** | Broadcast delivery specs. |
| **Loudness: off** | Keep the mix exactly as you set it. |

Pick a target and click **Bounce** — the session mixes down and lands exactly on spec, and the choice is **saved on the session**, so every re-bounce hits the same loudness. Peaks are always kept below −1 dBFS, so a loud target on quiet, spiky material gets as close as it safely can instead of distorting.

The bounced mix flows back to wherever the session is used, so your video timeline always hears the latest master.

---

## Turn a recording into a meeting brief

If you recorded a meeting or interview, you don't have to rewatch it. The **Meeting brief** panel turns a recording into a faithful, skimmable summary — a paragraph overview plus **action items** (with owners where stated), **decisions**, **chapters**, and **follow-ups**. It only summarizes what was actually said; it never invents content.

You can either:

- **Summarize the selected recording.** Select a recorded clip on the timeline and click **Summarize selected recording** — CoreReflex transcribes the audio and briefs it in one step. If there's no speech in the clip, it tells you rather than inventing a summary.
- **Paste a transcript.** If you already have a transcript, paste it in and click **Summarize**.

The brief appears right in the panel. On a **Pro** workspace, you can also save a brief to a contact in the [CRM](client-operations.md), seed tasks from its action items, and fire automations off it. For the full workflow, see [Meetings & AI Briefs](meetings-and-briefs.md).

---

## The real-time voice agent

CoreReflex also runs a **real-time voice agent** — a bot you can *talk to* over your microphone that talks back instantly, in a natural voice, and can take real actions mid-conversation. It uses one live voice core with multiple persona packs: an appointment setter, a lead qualifier, a sales assistant, and a support agent. Depending on the persona, it can book a real calendar, capture a CRM lead, answer from your knowledge base, or request a human handoff.

### What it is

The voice agent is powered by **Gemini Live on Vertex AI**. You speak; it listens, thinks, and replies out loud in a warm prebuilt voice — with **barge-in**, so you can interrupt it and it stops talking the moment you do, just like a real conversation. Behind the scenes, your audio only ever travels between your browser and CoreReflex; the connection to the AI is bridged on the server, so credentials never reach the page.

### Try it

1. Open the voice widget at **`/voice`** on your CoreReflex domain — for example, [corereflex.com/voice](https://corereflex.com/voice).
2. Pick a **Persona** in the dropdown, or link straight to one with `?persona=appointment-setter`, `?persona=lead-qualifier`, `?persona=sales-assistant`, or `?persona=support`.
3. To have it book into a real calendar, add the calendar's booking slug to the link: **`/voice?calendar=<your-calendar-slug>`** — the same slug your public [booking page](client-operations.md) uses. Without a calendar bound, booking personas will tell the caller the line isn't connected yet rather than booking nowhere.
4. Tap **Start talking** and allow microphone access. The orb lights up when the line is live and the status reads *"Listening — go ahead and speak."*
5. Speak naturally. Ask for an appointment and the booking-capable personas will offer **real open times** pulled live from the calendar, let you pick one, take your **name and email** (reading the email back to confirm the spelling), and **book the slot** — creating a real calendar event, a real booking, and a CRM contact, and firing your booking automations, exactly as if someone had booked through the web page. Lead and support personas capture the contact or handoff into the CRM instead.
6. Tap **Stop** when you're done.

### Choose the right persona

| Persona | Best for | What it can do |
|---------|----------|----------------|
| **Appointment Setter** | Friendly front-desk scheduling | Check availability and book the exact slot the caller chose. |
| **Lead Qualifier** | Inbound interest capture | Gather name, contact details, timeline, budget, and interest, then save the lead. |
| **Sales Assistant** | Qualified sales conversations | Answer grounded sales questions, capture the lead, request handoff, and book a call when appropriate. |
| **Support Agent** | First-line support | Search your knowledge base, answer from it, and request human follow-up when the answer is missing or sensitive. |

### How booking behaves

- **It only books — and never invents availability.** Every time it offers slots, it checks the live calendar first; it will only confirm a time that's genuinely open, and it reuses the exact slot you chose so the booking matches what you heard.
- **If a slot was just taken,** it apologizes briefly and offers another.
- **It stays on task.** Ask the appointment setter something off-topic and it gently steers back to booking. Sales and support personas can answer grounded questions through your knowledge base, but they still avoid making up facts.
- **It speaks in short, conversational turns** — this is a spoken assistant, not a chatbot reading paragraphs.

### Good to know

- The voice line is **rate-limited and time-bounded** to keep it safe and affordable: there's a cap on how many sessions can start from one place per minute, a global limit on simultaneous live conversations, and a per-session time limit plus an idle cutoff (if you go quiet for a while, the line politely hangs up). These are sensible defaults you won't normally bump into.
- Because the calendar tools are the **same booking engine** as the public booking page, anything the agent books shows up in your dashboard alongside web bookings — see [Client Operations](client-operations.md).
- Phone calls use the same voice core through the **Twilio telephony bridge**. When your operator connects a Twilio number to `POST /api/voice/twilio?calendar=<slug>&persona=<persona>`, callers reach the same Gemini Live agent over the phone, and support handoffs can create CRM activity from the caller's number.
- CRM transcript logging is available when the session is tied to a booking calendar and a contact can be resolved. Web and phone sessions record safe user/model/tool text into the contact activity timeline, so a lead or support handoff doesn't disappear after the call ends.

---

## The Chrome extension: quick tab capture

For grabbing a quick recording of a **browser tab** without opening the editor first, CoreReflex offers a lightweight Chrome extension, **CoreReflex Capture**. It records the current tab's video and audio and uploads the clip straight into your workspace, where it appears under your imports in the editor — the same place an in-editor recording lands.

The extension is currently a developer build you load yourself; it isn't on the Chrome Web Store yet.

**To install and use it:**

1. **Sign in first.** Open [corereflex.com](https://corereflex.com) and sign in to your account in the same Chrome profile. The extension uploads using your signed-in session.
2. In Chrome, go to `chrome://extensions`, turn on **Developer mode**, choose **Load unpacked**, and select the extension folder from the CoreReflex project.
3. Open the tab you want to record, click the extension icon, then **Record this tab**.
4. Click **Stop recording**, then open the editor — your clip will be under your imports.

A few notes:

- The extension records the **current tab only**. To capture a full screen, a window, your webcam, or your microphone, use the in-editor recorder above.
- Because it relies on your `corereflex.com` sign-in, make sure you're logged in before you record. If an upload fails, confirm you're still signed in and try again.
- The extension is a new build and **hasn't been load-tested in every Chrome version yet** — if you hit a snag, the in-editor recorder is the reliable path.

---

## Putting it together

A typical podcast episode, capture to shipped:

1. **Record** your screen, camera, or a tab — the clip lands on the **timeline** automatically.
2. **Caption** it, then tighten it **in the transcript**: click through the words, delete the flubs, hit **Remove fillers** and **Remove pauses**.
3. **Style** the captions with a one-click preset (and **translate** them if your audience isn't English-only).
4. **Master** it — **Podcast voice** for a quick polish, or the **Audio Studio** for a multitrack mix with **Auto-duck** and a **−16 LUFS podcast** loudness target.
5. **Render** the full episode, and **download the SRT/VTT** to upload alongside it.
6. If it was a conversation, generate a **[meeting brief](meetings-and-briefs.md)** so nothing gets lost.
7. Feed the episode to **[Auto-Clips](#auto-clips-turn-a-long-recording-into-vertical-shorts)** — ranked vertical shorts, rendered into your library and scheduled to your socials.

Everything stays inside your workspace, on your own fleet, ready to ship.

---

## Where to go next

- **[Editing & Mastering](editing-and-mastering.md)** — trim, layer, grade, caption, and master every recording.
- **[Meetings & AI Briefs](meetings-and-briefs.md)** — the full brief workflow, saving to contacts, and seeding tasks.
- **[Voiceover & Music](voiceover-and-music.md)** — add AI narration and an original score to anything you capture.
- **[Client Operations](client-operations.md)** — the booking calendars the voice agent books into, plus the CRM and automations.