Jounai.nl: Playing with New Tech to Reinvent the News
summary
Jounai is a live AI-driven news site that actually generates up-to-date articles, cites about 50 public trusted sources, and publishes multiple automated AI-generated podcasts on Spotify. The stack swaps AWS for Azure container apps, uses GitHub Actions, a Python AI engine with structured-output validation (instructor, Pandera), and a Java Spring Boot backend with Vue/Nuxt front end. Costs run roughly less than $2 a week for model API calls plus about $80/month for hosting. Hallucinations happen; added validation and simple Jaccard clustering make it usable but scaling raises cost and complexity.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:04AI-driven Junai platform
- 3:10Junai features
- 5:15Junai architecture
- 7:00Azure setup
- 8:25Java back-end
- 9:49Front-end journey
- 11:11AI news processing
- 13:19Hallucinations handling
- 14:25LLM output validation
- 15:06Speech generation
- 16:58No generated images
- 18:03Cost analysis
- 19:47Operational costs
- 20:30Future plans
- 21:42Question and answer session
related talks
01h 00mThe Future of Content: Context-Aware, Centralized, and Written by Robots
24mAI and Web Development: Hype or Reality
28mAI’s Next Great Challenge: How to create AI that truly understands us?
49mBuilding AI Enabled Web Apps
56mBuilding web apps with Cloud and AI
09mRethinking AI: Building What Matters, Not Just What's Possible
15mThe future of AI is now: Real-life case studies for on client-side AI adoption in web apps
06mQ&A with Jòan Varvenne