<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Aamir Orbit — Writing</title><description>Field notes on shipping AI systems that hold up in production — agent architecture and guardrails, retrieval quality, multi-tenant scaling, and how small teams close the gap between a demo and a deployment.</description><link>https://aamirorbit.com/</link><language>en</language><item><title>You Cannot Prompt Your Way to Safety</title><link>https://aamirorbit.com/writing/ai-agent-guardrails-action-layer/</link><guid isPermaLink="true">https://aamirorbit.com/writing/ai-agent-guardrails-action-layer/</guid><description>Web3 learned that you cannot secure an irreversible action with a warning message, because there is no undo. AI agents are acquiring the same property, and the industry is rediscovering the same four controls under new names.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><category>ai agents</category><category>security</category><category>guardrails</category><category>web3</category><author>contact@aamirorbit.com</author></item><item><title>Your First Engineering Hire Should Think Like a Forward Deployed Engineer</title><link>https://aamirorbit.com/writing/first-engineering-hire-forward-deployed-engineer/</link><guid isPermaLink="true">https://aamirorbit.com/writing/first-engineering-hire-forward-deployed-engineer/</guid><description>Forward deployed engineer job postings grew eightfold in a year. The role gets framed as an enterprise problem, but the mindset matters more at five people than at five hundred.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><category>hiring</category><category>forward deployed engineer</category><category>product engineering</category><category>startups</category><author>contact@aamirorbit.com</author></item><item><title>Failover That Loses State Isn&apos;t Failover</title><link>https://aamirorbit.com/writing/llm-provider-failover-stateful-routing/</link><guid isPermaLink="true">https://aamirorbit.com/writing/llm-provider-failover-stateful-routing/</guid><description>Two LLM providers at 99.53% each give you 99.99% combined. That arithmetic is real and it is also the easy half of the problem — because the conversation does not survive the switch.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><category>llm infrastructure</category><category>reliability</category><category>ai agents</category><category>architecture</category><author>contact@aamirorbit.com</author></item><item><title>What Breaks When Your AI Product Meets Its First Enterprise Customer</title><link>https://aamirorbit.com/writing/multi-tenant-ai-saas-first-enterprise-customer/</link><guid isPermaLink="true">https://aamirorbit.com/writing/multi-tenant-ai-saas-first-enterprise-customer/</guid><description>It is not the model. It is that a request stopped being a unit of cost, one customer can now exhaust another customer capacity, and someone is going to send you a security questionnaire.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><category>saas</category><category>architecture</category><category>scaling</category><category>llm infrastructure</category><author>contact@aamirorbit.com</author></item><item><title>Your RAG System Thinks It&apos;s Always Today</title><link>https://aamirorbit.com/writing/rag-recency-reranking-production/</link><guid isPermaLink="true">https://aamirorbit.com/writing/rag-recency-reranking-production/</guid><description>Embedding search has no opinion about time. It will hand you a policy that was superseded eighteen months ago with exactly the same confidence as the one that replaced it — and your users will not be able to tell.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate><category>rag</category><category>retrieval</category><category>llm infrastructure</category><category>search</category><author>contact@aamirorbit.com</author></item></channel></rss>