Point AI at Your Data And It Guesses. The Fix Is a Layer Most Teams Skip.
AI leader Kamal Singh says frontier models are nearly interchangeable. Enterprise AI reliability comes from the context, ontology and access control that surround them.
Data & Infrastructure
Data platforms, infrastructure, and the systems behind enterprise AI.
Postgres Teams Turn the Write Ahead Log Into a Real Time Event Bus
Diaa Radwan, Data and AI Solution Engineer on Microsoft's Global Black Belt team, explains how WAL, logical replication, and change data capture let PostgreSQL generate real-time events for downstream systems while staying the authoritative system of record.
Graph Queries In Postgres 19 Reach Production Only When Every Extension Version Aligns
Vladislav Galibov, Co-Founder of Croco.Team, explains why the extensions under a Postgres AI stack decide when a new release becomes usable in production.
AI is Telling Engineers to Build the Database First and the Rest Can Follow
Grant Fritchey, Product Advocate at Redgate Software, explains why AI-assisted development is pushing enterprises back toward data-first architecture, where database design, governance, and senior engineering judgment decide whether faster code holds up.
Governance and Lineage Help Data Leaders Reconcile Departmental Records for Production AI
Nandakumar Sivaraman, SVP and Chief Architect of Enterprise Data at Bridgenext, explains why AI needs to know who owns each copy of a record and how it changed.
KV Cache Optimization Turns Inference Into a Storage Systems Problem
Intel engineer Vamshi Nagireddy explains why KV cache optimization is a systems problem best solved with the block allocators, eviction policies, and data-locality discipline of databases.
Enterprise AI Stalls on Data Quality While Boards Keep Asking About the Model
Aaron Pries, Director of Data Analytics at Bluenet, explains why enterprise AI value depends on backend data quality, API connectivity, and system integration rather than client-facing features.
Software Architecture Fundamentals Matter More Now Than at Any Point in the Industry's History
Matt Erman, Senior Software Engineer at Blackboard and author of the forthcoming Grokking Software Architecture, argues that AI's coding speed is making architectural fundamentals, security discipline, and human accountability the real engineering moat.
Enterprises Don't Need Perfect Data To Start Preparing For AI Agents
Suggestic CEO Carver Anderson says enterprises can start building AI pipelines on messy data today, laying the groundwork for a future where brand agents replace the website.
Production Agentic AI Needs Deterministic Guardrails Not Another Chatbot Layer
Joseph Huffnagle, Global Vice President of Field Engineering and Delivery at Parloa, explains why enterprises moving agentic AI into production need agentic-first architecture with deterministic guardrails where regulation demands them, not scripted automation dressed up as autonomy.
Production AI Stalls When Teams Treat Retrieval Design and Cost Discipline as Afterthoughts
Trent Mcfarlane, Senior AI Architect at Zero Click Dev, argues that production AI fails when teams skip retrieval design, token cost discipline, and gateway infrastructure.