Khushi Shah·3d agoWhy Prompts Don’t Stop AI Agents From Leaking Secrets (And What Does)A practitioner’s take on agent identities, secret injection, and why the safest agent is one that never sees the key.
Khushi Shah·4d agoEnterprise AI Agent Governance: My Take on Why Ownership Beats PolicyA developer-facing walkthrough of AI agent governance: registries, scoped credentials, runtime policy checks, approvals, and audit logs.A response icon1A response icon1
Khushi Shah·Nov 6Can a Web Page Hijack Your AI Agent? My Take on Indirect Prompt InjectionYes, and the fix isn’t a smarter prompt. Here is why I think tool output should be treated like hostile input, and what actually helps.
Khushi Shah·Nov 5Why Your Small SaaS Cloud Bill Keeps Growing (and the 6 Fixes I’d Make First)A small SaaS cloud bill keeps growing because the decisions behind it are made fast, early, and almost never revisited. The fix is not a…
Khushi Shah·Nov 4Can You Prevent AI Hallucinations in a Customer-Facing Product? My TakeI don’t think you can eliminate them. I do think you can design a product where they’re rare, visible, and easy to recover from.
Khushi Shah·Nov 3Can AI Route Support Tickets? My Honest Take on What WorksAI is good at sorting tickets and bad at judgment. Here is where I think the line sits, and why more teams are building routing themselves.A response icon1A response icon1
Khushi Shah·Nov 2Can You Build a SaaS MVP Without Coding? My Honest Take for Non-Technical FoundersWhy more founders are skipping the developer hunt, where no-code and AI builders hold up, and where they still need a human.A response icon1A response icon1
Khushi Shah·Oct 30How to Build an AI Customer Support Agent That Knows When to Hand Off to a HumanMy take on why the handoff, not the answer, decides whether support automation earns customer trust.A response icon1A response icon1
Khushi Shah·Oct 29Can AI Write Technical Documentation From Code? My Take on What to TrustAI is good at extracting structure and drafting. It is not good at knowing why. Here is where I would trust it and where I would check.
Khushi Shah·Oct 28AI-Generated Wireframes: Are They Usable? My Take on What Makes Them WorkMaking a wireframe takes minutes now. Making one a team can decide with is still a human job. Here’s where I think the line sits.