14d • Open Chat
From self-doubt to delivering a full AI Receptionist built on Vapi here's what we built
A few months ago I was doubting whether I could
really make this work.
Lost projects. Slow months. Wondering if
I was cut out for this.
I kept going anyway. Kept building. Kept learning.
Last week I delivered a complete AI Receptionist
for Cinco Farm a wedding and event venue
in Miami, Florida.
Meet Giselle. 🌸
She's the AI receptionist I built on Vapi and
here's what she does:
1 Answers every call with:
"Thank you for calling Cinco Farm. This is
Giselle at the guest concierge line, how can
I help you plan your beautiful event today?"
2 Fully bilingual English & Spanish
Detects the caller's language in real time
and switches instantly. No delay.
3 Answers questions about:
- Pricing and packages
- Venue information and capacity
- Services and event types
- Tours and availability
3 Books venue tours through Google Calendar
4 Captures leads, call transcripts, and
summaries automatically
5 Sends internal email notifications via
Zapier + SMTP the moment a call ends
6 Voicemail fallback for complex requests
Tech stack:
Vapi (voice AI)
Claude Haiku 4.5 (AI model)
Deepgram Nova 2 (transcription)
11labs Jessica voice (warm, elegant)
Zapier (automation)
Google Calendar (booking)
SMTP (email notifications)
Average latency: ~1,400ms
Average cost: ~$0.13/min
Client was thrilled. System is live and
handling real calls right now.
Next step: automated SMS once Twilio A2P
registration is approved.
If you're building AI receptionists or
exploring Vapi for client projects,
drop a comment or DM me.
What's the most interesting use case
you've seen for voice AI so far?
0
0 comments
From self-doubt to delivering a full AI Receptionist built on Vapi here's what we built
skool.com/paid-ad-secrets
The #1 Fastest Growing Paid Ads Community Online ⚔️
Leaderboard (30-day)
1
+35553
2
+33043
3
+912
4
+594
5
+34
Powered by