Archive: 2026/07

July 28 posts

• LLM-Native Recommendation Architecture After Netflix GenRec Jul 31
• How MCP Servers for AI Agents Bridge Fragmented Data Jul 30
• Control Reasoning Effort LLM APIs in Production Jul 28
• BM25 vs Dense Retrieval and SPLADE for Production RAG Jul 27
• Shipping LLM Browser Agents Without Breaking Production Jul 26
• Claude Opus 5 Prompt Injection Hits 0% Across 129 Tests Jul 25
• AI Coding Agents Cut Per-Task Cost Up to 4x With Indexing Jul 24
• ChatGPT Shared Link Vulnerability Plants Rogue Agents Jul 23
• Grok vs Copilot in Excel for Builders Jul 22
• Run AI Models Locally on Mac With MLX and Nativ Jul 21
• Hugging Face Agentic Attack Redefines AI Security Response Jul 20
• AI Agent Cost Per Resolution Decides If It Ships Jul 19
• Self-Host LLM Inference, the Netflix Decision Framework Jul 18
• Where AI in Media Production Workflows Actually Pays Off Jul 17
• Orchestrate Open Source LLMs vs Frontier Models Jul 16
• OpenAI Codex Subagent Encryption Breaks Agent Observability Jul 15
• ChatGPT Sales Workflows Fail on Real CRM Data Jul 14
• Real-Time AI Dental Image Verification Cuts Claim Denials Jul 13
• Fine-Tuning vs RAG vs Prompt Engineering Decision Framework Jul 12
• How LLM Agent Scaffolding Fixes Failing Code Review Agents Jul 11
• Building Proactive AI Agents With Context Graphs Jul 10
• AI Coding Benchmarks Are Gameable by Design Jul 09
• AI Video Editor Limitations Break Iterative Workflows Jul 06
• LLM Vendor Data Risk Has a Break-Even Price Jul 05
• Stop RAG Hallucination with Typed Schema Contracts Jul 04
• Domain-Specific LLM Evaluation Demands Leakage-Free Data Jul 03
• Why an Agent-to-Agent Gateway Beats Point-to-Point Links Jul 02
• LLM API Token Inflation Hides Claude Sonnet's Real Cost Jul 01