Insights on AI assets, skills, and tools from the AI Khazna team.
AI answer engines now sit on top of nearly half of all searches. Here is a plain-English playbook for getting your pages quoted and cited inside ChatGPT, Gemini, Perplexity, and Google's AI Overviews.
Arabic is not one language but a formal standard plus a dozen living dialects — and AI handles each very differently. A practical 2026 guide to MSA vs dialect, the regional models worth knowing, and how to prompt for results you can actually use.
An MCP server is how an AI model reaches your tools and data through one common interface instead of dozens of custom integrations. A plain-English explainer of the protocol now supported by ChatGPT, Gemini, Claude, and Copilot alike.
Most poor AI results come from poor prompts, not weak models. A plain-English guide for business owners in the Gulf and beyond: the five parts of a strong prompt, the mistakes that quietly cost you, and how to get reliable output in Arabic and English.
AI search needs its own scorecard. A practical guide to the GEO KPIs that matter for brands across the GCC and wider MENA in 2026 — share of AI voice, citation rate, AI referral traffic, sentiment — and how to measure them in both Arabic and English.
Skills, MCP servers, agents, prompts, and workflows — five terms used interchangeably and rarely explained in Arabic. Here is how they actually fit together.
