Latest Articles: page 11
692 posts from Alex's blogs and newsletters, newest first. Thoughts on tech, data, policy, and philosophy.
Operational vs. Analytical Systems: Why the Oldest Divide in Data Exists, What Physics Enforces It, and the Honest Truth About Hybrid Systems
Every data architecture ever drawn contains the same fault line, so old and so universal that most engineers stop seeing it: on one side, the syste......
Personal Context vs. Shared Context: A Deep Dive Into How Humans and Organizations Should Feed Their AI Agents
The most important discovery of the agent era fits in one sentence: most AI failures are context failures, not model failures. When your assistant......
Designing Private, Air-Gapped Data Lakehouses: Scaling Iceberg in Highly Secure, On-Premises Clouds
Some of the most important lakehouse work happens in environments that will never look like a simple public-cloud reference architecture. Defense......
Migrating Proprietary Warehouses to Open Lakehouses: The 2026 Playbook for Zero-Copy Metadata Translation
Every warehouse migration sounds simpler before the first inventory. Then the team discovers old dashboards, hidden dependencies, undocumented stor......
Lakehouse Table Formats in 2026: Iceberg, Delta Lake, Hudi, Paimon, and DuckLake, How They Work, Where They Stand, and Where They're Going
The table format war is over, and the table formats are not. Both halves of that sentence are true, both matter, and the tension between them is ex......
Trustworthy Concurrency in the Agentic Lakehouse: Reconciling Academic Proofs with High-Frequency Production Writes
Agentic lakehouses change the concurrency conversation. Traditional data pipelines already deal with overlapping jobs, retries, compaction, merges......
When Gatekeepers Panic: The Encyclopédie, Open AI Models, and the Politics of Accessible Knowledge
The fight over open AI models mirrors the 18th-century suppression of Diderot's Encyclopédie, revealing the same pattern of institutional fear....
The Who, What, and Why of Semantic Layers: The Layer That Decides Whether Your Numbers Can Be Trusted
There is a survey statistic making the rounds this year that I cannot stop quoting: 84 percent of data teams report regularly encountering conflict......
Healthcare What the World Teaches Us: Diagnosing a Broken System and Charting a Practical Path Forward
TL;DR The United States spends $14,885 per person on healthcare annually, more than double the average of comparable wealthy countries, and ranks last...
Rules Without Rulers: The Case for Non-State Regulation
TL;DR Libertarianism is not about the absence of rules....
AI-Ready Metadata Prevents Query Failures
AI-ready metadata reduces query failures by making ownership, freshness, lineage, quality, and policy visible at execution time....
Autonomous Materialization for Agentic Analytics
Autonomous materialization is useful when it is tied to workload evidence, governance checks, and lifecycle management....
Composable Semantic Layers for Analytical Agents
AI agents need more than metric names. They need composable business logic that survives multi-step analysis....
Built for Agents and Managed by Agents
Dremio Agentic Lakehouse is easiest to understand as two ideas: data built for agent access and platform work managed by agents....
ClickHouse in the Loop for Active Agents
Low-latency analytical systems can help active agents, but only when event loops include validation, context, and safety boundaries....
The Context Layer for AI Agents
A semantic layer is necessary, but agents also need lineage, quality, freshness, compliance, and ownership context....
Lakehouse as the Operating Layer for Agentic AI
Agentic AI announcements are useful when they validate the need for governed data, semantic context, and cost-aware execution....
Event-Driven Table Compaction with Agents
Event-driven compaction is valuable when agents coordinate maintenance with workload signals, table health, and commit safety....
Fabric Agentic Analytics and Lakehouse Schema Design
Microsoft Fabric agentic analytics is a reminder that schemas, semantic models, and governed lakehouse design now shape AI behavior....
Fine-Grained Security for AI Agents
Machine-speed analytics requires machine-enforced policy, identity, masking, filtering, and audit controls....
Iceberg v4 Performance: Root Manifests and Calls
Apache Iceberg v4 discussion should focus on planning cost, metadata layout, and object storage round trips, not vague claims about faster tables....
What Is LTAP in the Lakehouse?
Lakehouse transactional analytical processing is useful only when teams define freshness, isolation, and workload boundaries clearly....
The Model Is Not the Moat
Enterprise AI advantage increasingly comes from governed context, semantic models, and operational data contracts, not only from model choice....
PyIceberg at Scale Without Apache Spark
Python-first Iceberg work is useful when it stays honest about what Python should and should not do....