Trustworthy AI in the Agentic Lakehouse: Reconciling Concurrency and Isolation Contracts
A single AI agent querying your lakehouse is manageable. A hundred AI agents : running automated monitoring, answering stakeholder…
The archive
Writing, page 23
593 posts on Apache Iceberg, lakehouse architecture, data engineering and applied AI.
The 2026 Unified Data Architecture: Reconciling Multi-Cloud Data Lakehouses
Three years ago, "multi-cloud strategy" for data meant maintaining separate warehouses on AWS, Azure, and GCP, then running ETL to sync…
Why Traditional Lakehouses Fail AI Agents: The Mathematical Case for the Agentic Lakehouse
When organizations first try connecting an LLM to their data lakehouse, the experience follows a predictable pattern: early demos work…
The Era of Zero-ETL Federation: Fueling AI Agents with Real-Time Cross-Enterprise Data
ETL pipelines were the right answer in 2010. You pulled data from operational systems nightly, transformed it, and loaded it into the…
Use Hermes Agent for Free With DeepSeek V4 and Slack
Most AI agent frameworks lock you into a paid model. Claude Code needs an Anthropic subscription. Codex needs an OpenAI plan. Cursor costs…
Concurrency, Isolation, and MVCC: How Engines Handle Contention
This is Part 10 of a 10-part series on query engine design. Part 9 covered distributed joins. This final article covers how engines handle…
Hash, Sort-Merge, Broadcast: How Distributed Joins Work
This is Part 9 of a 10-part series on query engine design. Part 8 covered partitioning. This article covers the most expensive operation in…
Partitioning, Sharding, and Data Distribution Strategies
This is Part 8 of a 10-part series on query engine design. Part 7 covered memory management. This article covers how engines divide data…
Buffer Pools, Caches, and the Memory Hierarchy
This is Part 7 of a 10-part series on query engine design. Part 6 covered execution models. This article covers how engines manage their…
Volcano, Vectorized, Compiled: How Engines Execute Your Query
This is Part 6 of a 10-part series on query engine design. Part 5 covered how optimizers pick a plan. This article covers what happens…