Lightning Talks - Day 1
summary
Five-minute slots report concrete claims and working tools: Devron demos a way to do “data science without data” and is hiring; trustor (Python 3.10 required) promises native OS trust store SSL contexts (experimental, do not deploy); Memray: sees C allocations, flame graphs, live reports (open‑sourced two weeks earlier); synthetic data case study raises accuracy from 88% to 96%; Holoviz demos rasterize=true for 11M points; a Rust‑based web runtime beats others on 10k requests; one project reaches 100% coverage in Jan 2022.
AI-written summary · about
chapters (13)
- 0:00Start
- 0:50Sameer Wagh - Data Science without Data?
- 2:25Cheuk Ting Ho - Cultural Shock - My 1st
- 7:25Łukasz Langa - COVARIANCE/CONTRAVARIENCE
- 11:45Seth M Larson - Truststore: OS trust stores in Python
- 15:50Pablo Galindo - Memray: hardcore memory profiling
- 20:17Graham Waters -The grief cycle, data security breaches, how we could code the
- 24:17Mason Egger - What is Synthetic Data
- 29:55Sophia Yang - Holoviz
- 34:00Shiray Lamba - Robyn; The fastest rust based python webframework server
- 39:00Chris May - Three steps to elegant code
- 43:50Chris Ariza - Getting to 100% coverage
- 49:15Indra - Jupyter ML model to production ML as a service
