Writing
Notes on building AI that ships
Agentic AI, architecture, scaling teams, and the fractional-CTO playbook — from the trenches.
Aug 16, 2026Engineering Trust in MENA's Risk-First CultureAug 16, 2026GPT-5.6 Sol Ultrafast: What 750 Tokens/s Changes for BuildersAug 15, 2026Agent Idempotency: The Production Gap That Breaks Multi-Agent PipelinesAug 15, 2026DeepSeek V4-Pro 4× Price Spike: Fix Your Unit Economics NowAug 14, 2026Voice AI Interrupt Handling: Fix the Latency Gap Killing ConversationsAug 14, 2026Grok 4.6: Frontier Intelligence at $2/M — and the 200K CliffAug 13, 2026Agent Handoff Contracts: The Protocol Gap Costing You Production ReliabilityAug 13, 2026Daybreak Blue/Red: API Access Just Became a Compliance DisciplineAug 12, 2026Where AI ROI Is Real: A Decision Framework for 2025Aug 12, 2026GPT-5.6-Cyber: Offense-Grade AI and What Gated Models MeanAug 11, 2026AI Roadmaps That Survive Technical Due DiligenceAug 11, 2026Muse Glimmer 30B: When a Consumer GPU Runs Your Agent StackAug 10, 2026Shadow Evals: The Test Layer That Catches Regressions Before Users DoAug 10, 2026GPT-5.4 Codex Deprecation: Aug 31 Migration ChecklistAug 9, 2026AI Cap Table Audit: 5 Vendor Traps Killing Your ValuationAug 9, 2026GPT-5.6 Sol Fast Mode: The 2.5× Speed Play and What Luna's 80% Cut RewritesAug 8, 2026Dual-Track Engineering: How MENA Teams Build for Two Markets at OnceAug 8, 2026Sol's Effort Slider: The Inference Cost Lever to Wire NowAug 7, 2026Streaming vs. Batch LLM Calls: The Latency-Cost Decision TableAug 7, 2026Gemini Deprecated temperature and top_p: Audit Every API Call NowAug 6, 2026Agent Context Windows: The Throughput Ceiling Nobody Budgets ForAug 6, 2026gpt-5.5 Is Live in the API: 25-Day Migration Deadline and New Routing MathAug 5, 2026AI Due Diligence: 5 Data Room Gaps That Kill Funding RoundsAug 5, 2026GPT-5.6 Luna 80% Cut: The Agent Routing Math Just ChangedAug 4, 2026Feature Flags for LLM Changes: Ship Fast, Roll Back in 30sAug 4, 2026DeepSeek V4-Flash-0731: Silent Upgrade, Real Risks to Audit NowAug 3, 2026Kubernetes Autoscaling for AI Workloads: Stop Paying for Idle GPUAug 3, 2026Gemini 3.6 Flash GA: Reprice Your Agentic Output Before FridayAug 2, 2026Agent Retry Logic: The Silent Cost Multiplier in Multi-Agent SystemsAug 2, 2026EU AI Act Enforcement Day Zero: What Builders Must Do NowAug 1, 2026GPT-5.6 Luna at $0.20: Reprice Your Agent Routing Stack NowJul 29, 2026MCP 2026-07-28: What the Stateless Shift Means for Your StackJul 28, 2026The Hidden Chunking Variable Breaking Your RAG PipelineJul 27, 2026Agent Trust Boundaries: The Security Layer Multi-Agent Systems SkipJul 27, 2026Kimi K3 Weights Drop: The Self-Host Decision Is Now a Compliance CallJul 26, 2026Multi-Tenant K8s: The Namespace Cost Leak Nobody AuditsJul 26, 2026Claude Opus 5: Cost Reset That Rewires Your Routing LogicJul 25, 2026The Fractional CTO Handoff: How to Make Yourself UnnecessaryJul 25, 2026Computer-Use vs. Tool Calling: The $1B Architectural Bet You Need to Resolve NowJul 24, 2026Kubernetes Cost Traps: 3 Misconfigs Draining Your Cloud BillJul 24, 2026OpenAI's Triple Deprecation: Migrate Your Evals and Agents Before October 31Jul 23, 2026The 3-Layer AI Readiness Stack: Data, Workflow, and DecisionJul 23, 2026OpenAI Presence: When Your Model Provider Becomes Your CompetitorJul 22, 2026Agent State Machines: Stop Runaway Agents With Explicit StatesJul 22, 2026Inkling: The US Open-Weights Model That Rewrites the Self-Host CalculusJul 21, 2026Prompt Caching Is the Fastest ROI in Your LLM Stack Right NowJul 21, 2026GPT-5.6 Context Window Cut: The 2× Billing Trap AuditJul 20, 2026Agent Observability: The Telemetry Gap Killing Your Agentic StackJul 20, 2026Agentic AI Just Ran Its First Confirmed Production BreachJul 19, 2026Tool Calling at Scale: Why Agent Tool Design BreaksJul 18, 2026Agent Failure Modes: 5 Production Breaks Nobody DemosJul 17, 2026RAG Reranking: The One Layer Most Teams Skip That Kills PrecisionJul 17, 2026Kimi K3: What 2.8 Trillion Parameters Actually Means for Your StackJul 16, 20265 Signals It's Time to Hire a Fractional CTOJul 15, 2026Hybrid RAG: When Vector Search Alone Stops Being EnoughJul 15, 2026ChatGPT Work Has Write Access: The Security Tax You Must PayJul 14, 2026Fine-Tune vs. Prompt Engineer vs. RAG: A Decision Table for 2025Jul 14, 2026Grok 4.5: The Token Efficiency Number That Rewrites Agentic Cost MathJul 13, 2026The MENA Engineering Leadership Traps No Playbook Warns You AboutJul 13, 2026Meta Muse Spark 1.1: The 6x Output Cost Gap Changes Inference Math
The AI CTO playbook
Get my AI playbooks — straight to your inbox
Practical notes on shipping production AI, scaling teams, and the calls a CTO actually has to make. A few times a month. No spam, no fluff.