Slide 1
archive.ire.org The next generation of investigative knowledge sharing
Slide 2
33,449 resources Now available in a new searchable database
Slide 3
What you'll find Thousands of contest entries, tipsheets and training materials
Slide 4
What we've got Some fields are better than others
Slide 5
Semantic search AI helps find matches based on the meaning of your words
Slide 6
Tune the results Controls allow you to filter and sort
Slide 7
Discover new things AI recommends related documents
Slide 8
Members only Restricted to active members with an ire.org account
Slide 9
The software stack Vercel frontend SvelteKit: Static single-page application consumes JSON data Fly.io backend FastAPI: Python application supplies JSON data Redis: Caching layer enforces ire.org membership wall Qdrant: Vector database indexes with semantic embeddings
Slide 10
Work in progress Report problems to help@ire.org
Slide 11
What's next 1. Create staff tools: Empower IRE staffers to update the database 2. Complete the migration: Upload the 10% of resources that aren't indexed yet 3. Tidy up: Reconcile imperfections in the metadata 4. Add audio transcripts: Integrate text from thousands of recordings transcribed by AI
Slide 12
Your humble servants Tell us what stinks so we can make it better Ben Welsh: b@palewi.re Derek Willis: dwillis@gmail.com Cody Winchester: cody@ire.org