Reaching One Million Hands for Israeli advocacy
summary
One Million Hands starts as automated Google Sheets scripts generated with GPT and evolves into a cloud platform with a Postgres-backed data pipeline, scrapers and ML models that tag violence, predict virality and rank posts. Over 600 volunteers average 1.5 hours/day, producing about 45,000 posts/day and 0.5–2.5 million daily reach (13–60M/month). The initial Flow app collapses under news-driven traffic, so the group rebuilds a scalable end-to-end system and measures impact with SQL analytics.
AI-written summary · about
chapters (8)
- 0:00Introduction to Million Hands
- 3:50Social media advocacy challenges
- 7:50Initial solution and Dirt app
- 10:25Transition to professional platform
- 13:19Data architecture and development
- 16:11Data science insights
- 18:45Python's role in project
- 20:32Thanks and closing
related talks
43mBehind the Scenes at Gapminder: A Story about Data and Visualizations
36mML: Predictive analytics in Ad Engineering
30mFrom Pixels to Plots: Transforming Lives of African Smallholder Farmers through Geospatial Data
34mAyudaPy org: From Weekend Project to Civic Movement
29mGiant Pile of Data
16mThe importance of representative data
41mUsing web data for a better tomorrow
11mData science: Next-gen performance analytics