- Health metrics (performance, bottlenecks)
- Metadata: tables size, num (min, max, ...), string (distinct, common)
- Costs: which models are expensive, why real time?
- Dashboard for monitoring
- Query annotations
- Data validation + anomalies
- Yara Fertilizers at.farm
- Ourworldindata.org
- Correct ratio of data engineers vs scientists (2-1)
- DataEng democratizes data
- tiny.dbi.io/detbook
- Dataflow simplifies? Easier but not simple
- Jupiter: AAS with branches (no data stored)
- Vars as ENV
- Bigtable R/W speed
- Redis secondary
- Nowcasting vs. forecasting
- word2vec with journeys
- Taxis 30% use time vs. Cabify 55%
- Tracking + domain events
- Realtime patterns: event + wait time = action
- Flink + Airflow
- Precomputed tables