Building fun and creative messaging experiences on WhatsApp
summary
Build turns selfies into anime-style images using Twilio, OpenAI image generation APIs, sharp for deterministic overlays, ngrok for local webhooks and the WhatsApp sandbox for testing. Measured results: 98% open rate for messages, ~30% click-through and Twilio reports 6 billion business messages/day; demo handles bursts (one attendee uploaded ~55 images) and the chosen OpenAI model 1.5 plus sharp keeps overlays stable and latency reasonable in live demos.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 2:35Why communication matters
- 7:49Messaging evolution history
- 11:20Business messaging vs email
- 16:11Messaging adoption trends
- 20:00RCS versus SMS
- 22:47Branded messaging importance
- 26:18Fun engaging messaging
- 29:04Live demo setup
- 33:16WhatsApp image generator bot
- 43:04Choosing AI models
- 47:59Deploying with Twilio
- 55:21WhatsApp bot architecture
- 58:52Resources and next steps
related talks
01h 01mBuilding fun and creative messaging experiences on WhatsApp
59mHow to Build a Quokkabot
28mScrew the Browser! Build ChatGPT via SMS
01h 20mKeynote
26mHow to enable personalised experiences using Conversational.
08mThe Architecture of Meta Business Agent on WhatsApp
26mF8 Refresh | Innovation Showcase: Meet some of the brightest startups building with Facebook
20mWhat's Up WhatsApp? - Two Years After the Big Deal