Swift Chatbots for Fun! and Profit
summary
Builds chatbots as server-side Swift REST services using Vapor + Heroku, exposing a callback endpoint that receives Messenger JSON and replies via the Messenger webhook or push endpoints. Shows concrete examples: an echo bot, a translator that proxies translation APIs, an OCR-enabled vocabulary helper, and a scheduled xkcd fetcher using the xkcd API with persisted metadata for search. Recommends solving one or two tasks well, and delegates heavy NLP to Python microservices or cloud APIs, saying robust NLP is ~2–3 years away.
AI-written summary · about
chapters (17)
- 0:00Intro
- 0:46Chatbot craze
- 1:06Chatbot history
- 1:50Why Swift
- 2:34Things to consider
- 3:01What is Chatbot
- 4:41Behind the Scenes
- 5:59Endpoints
- 6:26Custom GUI
- 6:56Vapor
- 7:37Route Swift
- 8:58JSON
- 9:25XKCD
- 9:52API endpoints
- 10:26What is a good chatbot
- 12:53Natural language processing frameworks
- 14:11Outro
related talks
10mSwift Chatbots for Fun! and Profit
06mSwift in my home
39mFull Stack Swift Development
57mSmart Chat App with Vue, AI and Serverless Function Solution LIVE
45mDr. Strangelove or How I Learned to Stop Worrying and Love the Serverless Chatbots
20mLive coding a streaming ChatGPT proxy with Swift OpenAPI—from scratch!
39mBYOB: Build Your Own (Swift) Backend. Server-Side Swift beyond CloudKit
10mServer-Side Swift: Surviving or Thriving?
