Measuring Success - Observability and More!
summary
Tracksuit sets up concrete observability: DBT + Elementary on Snowflake to quantify data health with defined formulas for accuracy (accurate records/total), reliability (incidents or test failures over time), and completeness (non‑null ratio). The deployment shows 435 models, 84 monitored (~19%) with few failures but recurring warnings surfaced in Elementary; median model run times feed a time‑to‑deploy signal and flag costly 20‑minute materializations. They also track analytics turnaround, self‑service vs ad‑hoc ratio, onboarding time and incident rates to reduce tech debt and firefighting.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 3:55Visible vs invisible work
- 7:52Technical debt and firefighting
- 12:33Observability for data quality
- 15:45Team productivity metrics
- 21:28Scaling metrics
- 24:40Truxt journey and future
related talks
31mNo-Nonsense Observability Improvement
42mObservability Is Not Analytics!
37mBetter Reliability Through Observability and Experimentation
13mObservability Where are We Headed?
47mObservability - Holistic Quality across Software Systems
35mObservability: Building Confidence & Answering the What, Where, How
18mTo observe or Not, is not the question
28mScaling Observability for Real-Time Personalization Engines