data-quality

Most wrong experiment results are not statistical subtleties; they are plumbing. This lesson covers how assignment actually works, hashing, not coin flips, why exposure must be logged at the moment of treatment, and the sample ratio mismatch check: the humble comparison of observed to expected group sizes that catches more broken experiments than any other single test.

Generation is the cheap half. What you discard determines what the student learns. This lesson orders the filters by strength: machine verification where an answer can be checked, self-consistency where it cannot, LLM-as-judge with its known position and length biases, and cheap heuristics. It ends on contamination, the failure that invalidates results rather than degrading them.

Look-ahead bias, survivorship bias, and point-in-time data. The errors that make a backtest wrong as a simulation, independent of any statistical question about whether the edge is real.

Most CRM projects fail, and almost never because of the software. This lesson covers the real reasons: poor user adoption, bad data, and weak change management, why the people-and-process problem dominates, how to design a CRM people actually use, and how CRM skill translates into a career.

Data governance is one of the most misunderstood functions in business: dismissed as bureaucracy, confused with IT or privacy law, rarely explained clearly. Learn what it actually is (managing data as a business asset through accountability and decision rights), the real cost of not doing it, why most programs fail as bureaucratic theater, and what the working version looks like.

Governance policy becomes real through concrete machinery. Learn the working parts every data governance program relies on: the dimensions that define data quality and how to measure them, metadata and catalogs that make data findable, lineage that traces where data came from, master data management that creates a single source of truth, and data contracts that push quality upstream.

The model is rarely the bottleneck. This lesson examines why data readiness — quality, governance, lineage, and access — is the primary constraint on enterprise AI value, with a practical scorecard, and a clear-eyed comparison of RAG versus fine-tuning economics.