Point of view · Oct 2026
How to Vet a Senior Data Engineer
Interviews test the talk, not the work. How to vet a senior data engineer with a practitioner-led screen, a paid trial task and a check of…
Read →
Skip to main content
Articles · points of view · data hacks
What we have learned, written down.
Points of view for leaders, practical data hacks for practitioners, and conversations from our webinars and podcast.
Point of view · Oct 2026
Nearshore or Offshore for Data and AI Work
For data and AI work, overlapping hours often matter more than the rate. When nearshore fits, when offshore works, and what to ask any provider before…
Read →Point of view · Oct 2026
Machine Learning or Generative AI?
Machine learning or generative AI? Start with the output: numbers, scores and plans point one way, language the other. How to choose, and when to use…
Read →Point of view · Oct 2026
MLOps in Practice
MLOps is a routine, not a platform purchase: a model inventory with owners, shared monitoring, retraining and retirement rules, and a review cadence.
Read →Point of view · Oct 2026
Model Drift: Monitoring and Retraining
Models decay quietly. What to monitor, from data quality to the business decision, and when to retrain a model on a trigger rather than a calendar.
Read →Point of view · Oct 2026
Measuring Literacy by Use, Not Attendance
Completion rates show a course ran, not that anyone works differently. How to measure AI and data literacy by use, quality and judgment, role by role.
Read →Point of view · Oct 2026
Data Literacy and AI Literacy
Data literacy and AI literacy are different skills. Why to score them separately by role, what each role needs, and how to measure use rather than…
Read →Point of view · Oct 2026
Agents and Credentials: Least Privilege
Treat an AI agent as a new kind of user: its own identity, the narrowest access, expiring credentials, an owner, and human approval for irreversible actions.
Read →Point of view · Oct 2026
What AI Adds to Your Attack Surface
The eight boundaries an AI request crosses, from prompt injection to agent credentials, and how to find, rank and close the exposures your program misses.
Read →Point of view · Oct 2026
Self-Service Analytics Without the Chaos
Self-service works when people explore governed measures, not raw tables. How a semantic layer, content levels and clear owners keep the numbers consistent.
Read →Point of view · Oct 2026
Why Dashboards Go Unused
Dashboards fail on trust and purpose, not design. Why reports go unopened, and how shared definitions, owners and retiring old reports bring readers back.
Read →Point of view · Oct 2026
NIST AI RMF in Practice
How to put the NIST AI RMF to work: named owners for each AI system, the evidence worth keeping, and a tiered review path that decides…
Read →