`personas/scripts/voice-cloning-onboarding.txt`

180 lines
================================================================================
CRANKD — VOICE CLONING ONBOARDING SCRIPT
For: new employees and contractors providing voice samples for ElevenLabs PVC
================================================================================

INSTRUCTIONS FOR THE SPEAKER
- Record in a quiet room. No fans, no AC hum, no music, no second voice.
- Use a USB condenser mic or your iPhone Voice Memos app held ~6 inches away.
- Speak naturally. Don't perform. Don't read like a newscaster.
- If you misspeak, pause 2 seconds and re-read that sentence — don't restart.
- Total target length: 30 to 45 minutes of clean audio.
- Save the file as WAV or high-bitrate MP3 (192 kbps or higher), MONO.
- Hand the file to Steve. Do not upload it anywhere yourself.

DO NOT SKIP SECTIONS. ElevenLabs needs phonetic variety to make a good clone.

================================================================================
SECTION 1 — ELEVENLABS VERIFICATION STATEMENT  (required, read first, verbatim)
================================================================================

NOTE: ElevenLabs occasionally updates the exact wording of their PVC consent
statement. Before recording, log in to elevenlabs.io, start the Professional
Voice Cloning submission flow, and COPY THE CURRENT STATEMENT VERBATIM from
their dashboard. The template below is the canonical shape — use it only if
ElevenLabs' UI is down or as a sanity check against their current text.

  "I, [SPEAKER FULL LEGAL NAME], confirm that the voice in this recording is
   my own. I am over the age of eighteen. I am providing these voice samples
   to ElevenLabs for the purpose of creating a Professional Voice Clone of my
   voice. I have read, understood, and agreed to the ElevenLabs Terms of
   Service and Privacy Policy. Today's date is [SPEAK THE DATE: MONTH, DAY,
   YEAR]."

Pause for two seconds before continuing.

================================================================================
SECTION 2 — PHONETIC RANGE PASSAGE  (≈ 3 minutes)
================================================================================

Read each paragraph at a natural conversational pace. Pause briefly between
paragraphs.

The rainbow is a division of white light into many beautiful colors. These
take the shape of a long round arch, with its path high above, and its two
ends apparently beyond the horizon. There is, according to legend, a boiling
pot of gold at one end. People look, but no one ever finds it.

When the sunlight strikes raindrops in the air, they act as a prism and form
a rainbow. The rainbow is a division of white light into many beautiful
colors. Throughout the centuries people have explained the rainbow in
various ways. Some have accepted it as a miracle without physical
explanation.

To the Hebrews it was a token that there would be no more universal floods.
The Greeks used to imagine that it was a sign from the gods to foretell war
or heavy rain. The Norsemen considered the rainbow as a bridge over which
the gods passed from earth to their home in the sky.

Others have tried to explain the phenomenon physically. Aristotle thought
that the rainbow was caused by reflection of the sun's rays by the rain.
Since then physicists have found that it is not reflection, but refraction
by the raindrops, which causes the rainbows.

================================================================================
SECTION 3 — NUMBERS, DATES, NAMES  (≈ 2 minutes)
================================================================================

Read clearly. These train the clone on tricky pronunciation.

The meeting is scheduled for Tuesday, January twenty-second, twenty
twenty-six, at three forty-five in the afternoon. Please confirm by replying
to this message or calling six-one-nine, five-five-five, zero-one-four-two.

The invoice total is one thousand, two hundred and forty-seven dollars and
ninety-three cents. The order shipped on March fifteenth and should arrive
by April second.

My name is Jill. I live in Sherman Oaks, California. I work with Steve
Abrams at Designer Wallcoverings. I can be reached at jill at example dot
com, or by phone at the number on file.

Common words people get wrong: schedule, almond, caramel, Worcestershire,
quinoa, route, pecan, syrup, salmon, often, espresso, sherbet, library,
Wednesday, February, candidate, mischievous.

================================================================================
SECTION 4 — EMOTIONAL & PACING RANGE  (≈ 4 minutes)
================================================================================

Read each line in the indicated tone. Don't overdo it. Subtle is better.

(Calm and reassuring)
Hi there. I just wanted to call and let you know everything is fine. There
is nothing to worry about. Take your time and call me back whenever you can.

(Mildly excited, smiling while you talk)
You are not going to believe this, but I just heard the best news. I had to
call you the second I found out. Are you sitting down? Okay, here it goes.

(Apologetic, slower)
I am so sorry to bother you with this. I know you are busy, and I know this
is probably the last thing you want to hear right now. But I think you
should know.

(Confidently delivering information)
The order has been confirmed. Your sample will ship today, and you should
receive a tracking number by end of business. If anything changes, I will
let you know right away.

(Casually asking a question)
Hey, quick question for you. Do you remember if we said three o'clock or
three thirty? I just want to make sure I have it right before I head out.

(Slightly tired, but warm)
It has been a long day, but I wanted to say thank you. Honestly. I really
appreciate everything you did this week. Get some rest.

(Pretending to read off a list, neutral)
Apples, oranges, bananas, pears, peaches, blueberries, raspberries,
strawberries, watermelon, cantaloupe, grapes, plums, mangoes, pineapples,
kiwis.

(Quick, energetic — like leaving a fast voicemail)
Hi, it's Jill. Sorry I missed you. Give me a call back when you get a
chance. Talk soon. Bye.

================================================================================
SECTION 5 — FREE-FORM CONVERSATIONAL  (≈ 5 minutes)
================================================================================

Improvise. Stay in your natural voice. Suggested topics — pick three and
talk for about ninety seconds each:

  - Describe your favorite meal and how you would explain it to someone who
    has never had it.
  - Tell the story of how you got your first job.
  - Describe the route from your front door to the nearest coffee shop, as
    if you were giving directions to a friend.
  - Talk about something you learned recently that surprised you.
  - Describe a movie or show you watched last week, without spoilers.
  - Explain to a ten-year-old how Wi-Fi works.

Speak naturally. Use "um" and "you know" if that's how you really talk —
the clone will sound more human for it. Do not perform.

================================================================================
SECTION 6 — CRANKD-SPECIFIC LINES  (≈ 2 minutes)
================================================================================

These are sample lines the clone is most likely to need to say. Read them
once each in a natural, friendly tone.

"Heads up — this is an A I generated voice call from Steve. You opted in
to receive this. Press one to continue, two to stop these forever, or
hang up at any time."

"Hey, sorry to bug you. I know you said you might need a fake reason to
leave the meeting — that's me. Just nod a few times and say 'okay, I'll
be right there'."

"Hi, this is Jill calling. Steve asked me to give you a quick wake-up
call. You said seven-thirty. It is now seven-thirty. Time to get up."

"This is a friendly reminder that your dinner reservation is in fifteen
minutes. Have a great time tonight."

"You asked me to call you in five minutes if you needed an excuse to
leave. This is that call. You're welcome. Bye."

================================================================================
END OF SCRIPT — TOTAL TARGET 30 TO 45 MINUTES
================================================================================

After recording:
  1. Save as WAV or 192+ kbps MP3, mono.
  2. Listen back once to confirm no background noise / dropouts.
  3. Send the file to Steve via the agreed channel.
  4. Do NOT upload to ElevenLabs yourself. Steve handles submission so the
     consent paperwork stays attached to the cloned voice.