Claire books around thirty sessions a month. After every booking, she writes a “here is what to expect” email. The structure is usually the same, but the names, locations, session types, and client details are different every time. She knows what she wants to say, but she spends too long finding the words.

This is not a chatbot. It is not a template with mail merge fields. It is a prompted AI workflow built around Claire’s own voice, so five minutes of input can become a draft she actually wants to send.

What happens next

Claire fills in five fields after a booking, and a draft that sounds like her appears in Gmail.

She reviews the email in a few minutes instead of writing it from a blank page.

Every client receives a message that feels personal, specific, and considered.

The AI never sends anything on its own. Claire reads the draft, adjusts it if needed, and sends it herself.

Four simple tools can return hours of writing time.

The tools are simple. The real work is in the prompt: documenting Claire’s voice, building the email structure, and making sure the draft sounds like something she would actually send.

Session intake

Jotform

Claire fills in five fields after confirming a booking. She adds the client names, session type, location, date, and a few notes from the inquiry call. That submission starts the workflow.

Connection layer

Zapier

Zapier receives the form submission and passes the details into the AI prompt. When the response comes back, Zapier creates a pre-filled Gmail draft in Claire’s inbox.

The writing

Claude API

The AI step uses the session details and Claire’s voice skeleton to create a full draft in her tone, her structure, and her sign-off. It gives her a strong starting point without making the email feel generic.

Review and send

Gmail

The draft lands in Claire’s Gmail. She reads it, adjusts anything she wants to change, and sends it. The final step stays with the person who owns the client relationship.

Why not use a template? A mail merge template can fill in names and dates, but the writing still sounds like a template. The AI step allows the sentence structure, tone, and details to change based on the session. A vineyard wedding in October should not read the same way as a beach family session in July.

Thirty sessions a month at thirty minutes of writing each is fifteen hours. Even if this tool cuts that in half, Claire gets meaningful time back every month. The API cost per draft is usually very small; the real work is in the setup, voice documentation, and prompt testing. For a business sending this many repeat-but-personal emails, the time savings can add up quickly.

A confirmed booking becomes a reviewed draft.

Click through each step to see what happens after Claire confirms a session booking. The workflow shows what she types, what the AI does with it, and what lands in her inbox.

Before the workflow

Claire fills in five fields, and that submission starts the workflow.

After confirming a family session booking with the Tremblays, Claire opens the form on her phone. She adds the client names, session type, location, date, and a quick note about what they talked about on the call.

The form is not a client questionnaire. It is a structured way of capturing what Claire already knows after every inquiry call, without asking her to write the full email from scratch.

Why this matters

The intake form front-loads the thinking Claire was going to do anyway. By the time she hits submit, the AI has the details it needs to write a specific and personal email. The form makes sure the input is clear before the draft is created.

Trigger Claire submits the form. Zapier receives the data and immediately passes it to the Claude API with her voice prompt.
Claire Leblanc Photography
Session follow-up intake · Internal use only
Client first name(s)
Sophie & Marc
Last name
Tremblay
Client email
stremblay@notarealemail.com
Session type
Family session — outdoor, lifestyle
Date
September 20, 2026
Time
5:30 pm (golden hour)
Location
Reservoir Park, Wolfville NS
Who is coming
2 adults, 2 kids (ages 4 and 7), family dog
Notes from the call what they mentioned, what matters to them
Sophie mentioned that the dog is a big part of the family and she wants him in lots of shots. Marc is a bit camera shy. The kids are energetic. They want the session to feel relaxed, not posed. Their anniversary is coming up, and these will be used for a canvas print.
Vibe one word or phrase
Warm, relaxed, real

Step 1 of 4

Zapier sends the variables to Claude, and Claude writes the draft.

Within seconds of Claire hitting submit, Zapier takes the form data and drops it into a prompt that already contains Claire’s voice skeleton. The prompt includes her tone, her structure, what she usually includes, and what she avoids saying.

Claude reads the variables, reads the skeleton, and produces a draft. This is the AI step. It is not a template with fields swapped in. It is a piece of writing that uses the session details the way Claire would use them herself.

Why this matters

The difference between a mail merge and this workflow is the difference between a generic reminder and an email that mentions Marc being camera shy, the dog, the anniversary, and golden hour at Reservoir Park. The AI reads the notes and writes from them, which is what a template cannot do.

Trigger Zapier receives the Jotform submission, constructs the full prompt with Claire’s variables inserted, calls the Claude API, and receives the draft response. The full round trip usually takes only a few seconds.
Claire’s writing workflow · Running
J
Trigger
New Jotform submission
Sophie & Marc Tremblay · Family session · Sep 20 · Reservoir Park
AI
AI Action
Claude API creates the draft email
Zapier sends the session variables and Claire’s voice prompt to Claude. Claude returns a full email draft.
What Zapier sends to Claude
Client: Sophie & Marc Tremblay
Session: Family lifestyle, Sep 20, 5:30 pm
Location: Reservoir Park, Wolfville NS
Who: 2 adults, 2 kids (4 and 7), and the family dog
Notes: The dog matters, Marc is camera shy, the session should feel relaxed, and the images are for an anniversary canvas.
Vibe: Warm, relaxed, real
Claire’s full voice skeleton is also included in the prompt.
G
Action
Gmail draft is created
The draft is created in Claire’s Gmail. It is addressed to Sophie, the subject is pre-filled, and the email is ready for review.

Step 2 of 4

The draft is waiting in Gmail, so Claire can review it before sending.

Claire opens Gmail. The draft is there, addressed to Sophie, with the subject line written and the full email ready. She reads through it, adjusts one sentence about Marc being camera shy so it sounds even more like her, and sends it.

The AI does not replace Claire’s judgment. It removes the blank page.

Why this matters

The blank page is where the time goes. Claire already knows what she wants to say, but she has to find the words from scratch thirty times a month. The draft gives her something to react to instead of something to create. Editing is faster than writing, and a good draft makes the editing easy.

Trigger Zapier creates the Gmail draft automatically after the Claude response comes back. Claire opens it in her normal Gmail inbox, so she does not need to learn a new tool.

Step 3 of 4

Sophie receives an email that feels like Claire wrote it, because Claire reviewed it before sending.

Sophie gets the email. It mentions the dog, the anniversary, Marc being camera shy, and golden hour at Reservoir Park. It reads like a photographer who actually listened on the call, because the notes from the call are what shaped the draft.

Sophie does not need to know that AI helped create the first version. What matters is that the email feels specific, thoughtful, and genuinely useful before the session.

Why this matters

The best client relationships begin before the session. An email that references specific details builds more trust than a polished but generic message. Claire was already taking those notes. This workflow makes sure the notes end up in the client experience.

Trigger Claire reviewed the draft, made a small edit, and sent the email from her Gmail account.

That is the full workflow. One form, one AI call, one draft, and one human review can turn a confirmed booking into a thoughtful client email. Let’s talk.

Step 4 of 4

The voice skeleton is what makes the workflow useful.

The Zapier setup is the easier part. The API connection is straightforward once the workflow is mapped. The part that makes the output sound like Claire, instead of sounding like a generic AI draft, is the prompt.

We document how Claire writes. We capture what she usually includes, what she avoids saying, how she opens, how she closes, and how she handles the details that matter to each client. That becomes the skeleton the AI writes into every time.

Why this matters

Anyone can connect Jotform to Zapier to an AI model and get a generic result. The voice documentation is what makes the output usable. The value is not just the wiring. The value is in capturing how a specific person communicates clearly enough that the workflow can support her without flattening her voice.

What we built We built a documented voice guide for Claire, a structured prompt that uses it, and a Zapier workflow that passes session variables into it. The prompt on the right is a simplified version of what runs inside every API call.
claire-voice-prompt.txt — sent to Claude API on every call
# ROLE You are writing a pre-session email on behalf of Claire Leblanc, a portrait and family photographer based in the Annapolis Valley. Write in her voice, using the session variables and notes below.
# SESSION VARIABLES Client names: {{client_names}} Client email: {{client_email}} Session type: {{session_type}} Date: {{session_date}} Time: {{session_time}} Location: {{location}} Who is coming: {{who_is_coming}} Notes from call: {{call_notes}} Vibe: {{vibe}}
# CLAIRE’S VOICE - Claire writes warmly, but she does not gush. - She sounds conversational and calm. - She references specific details from the call. - She does not use corporate language. - She does not say “amazing,” “stunning,” “perfect,” or “journey.” - She signs off as “— Claire.” - She uses one clear logistics block and then moves on.
# EMAIL STRUCTURE 1. Open by name and reference something specific about the session. 2. Include a practical details block with the date, time, location, and access note. 3. Use one or two paragraphs to speak to the details from the call. 4. Mention gallery delivery and any relevant print notes. 5. Invite the client to reply with questions and sign off as Claire.
# OUTPUT Return the email body only. Do not include a subject line, preamble, or explanation.

A marketing studio for small businesses trying to show up with the same care on the outside that they bring to the work itself.