⚡ Meta Global AI Developer Hackathon · Startup Track · Build for Your World

AutoAI Lead — 24/7 Bilingual AI Receptionist & Visual Demo Engine

Production extension for AutoAI Lead that turns any local business on Google Maps into an interactive mobile demo site with an English/Spanish voice receptionist powered by four chained Meta Model API endpoints.

🌐 AutoAI Lead Production (autoailead.com) ↗ 🦷 Live Sample Demo Site (/d/smile-dental) ↗ 💻 GitHub Repository ↗
4-MODEL META API PIPELINE ARCHITECTURE (https://api.meta.ai/v1) Content-Addressed SHA-256 Cache ($0.00 Repeat Runs) + Hard $5.00 Spend Cap
1. muse-voice-transcribe-1.0 POST /v1/asr/transcribe · 24kHz PCM WAV · ENDPOINTING + EN/ES languageBias ($0.18/hr)
2. muse-spark-1.3 POST /v1/chat/completions · Function Tool Calling + Strict JSON Schema ($1.25/$4.25 per 1M)
3. sam-3.1 POST /v1/responses (SSE) · Zero-shot object boxes + 1-bit pixel masks ($2.50/1k imgs)
4. muse-image-1.0 POST /v1/images/generations · On-brand WebP Hero Visual generation ($0.01/image flat)
Meta API Credit Pool
$150.00
Promotional Hackathon Credits
Hard Code Spend Cap
$5.00
Enforced Before Every API Request
Cumulative Real Spend
$0.0692
13 live API calls recorded
Zero-Cost Cache Hits
0
SHA-256 content-addressed replay
Instant Verified Scenarios ($0.00 Cache Replay):
Pipeline Controls
api.meta.ai/v1
Using pre-verified 24kHz PCM WAV recording
Spend Breakdown by Model:
1. Speech-to-Text
muse-voice-transcribe-1.0
Caller Audio Recording (24kHz PCM WAV) CACHED
Loading transcript...
Latency: — Audio: — Cost: —
2. Agentic Receptionist
muse-spark-1.3
Tool Calls + Strict JSON Schema CACHED
Loading receptionist response...
Urgency / Language—
Booked Slot—
Service Requested—
Est. Lead Value—
WhatsApp Owner Alert: —
Latency: — Tokens: — Tools: —
3. Scene Segmentation
sam-3.1
Prompt: "dental chair" CACHED
Business Photo
Latency: — Matched Objects: — Mask: —
4. On-Brand Hero Visual
muse-image-1.0
AutoAI Lead Mobile Demo Hero (WebP) CACHED
Generated Hero Visual
Latency: — Tokens: — Flat Rate: $0.01/img
Inspect Raw JSON Payload from /api/pipeline/run