The problems we solve for firms of ten to a hundred people, written down: security,
privacy, keeping an eye on what the assistant does, and what the standard lists mean
for a firm with no one whose job is AI.
Zero Mission Guide is in the App Store, and the latest version answers questions from a folder of your own documents, using the AI built into your iPhone. Nothing leaves the phone, there's no account, and it's free. We use it ourselves to test what makes answers better and faster.
AI can sound right and be wrong, and that's why people don't trust it at work. The fix isn't a smarter model. It's labeling: keep your firm's approved answers word for word and clearly marked, keep the AI's words separate, and the only mistake left is a wrong pick you can see.
The hard part of putting AI to work in a firm isn't the AI. It's keeping the firm's knowledge organized, current and seen only by the right people. Librarians have done that job for a century and a half, and it's the job AI needs done.
Locking everything down until a contractor asks keeps client records safe. It also leaves new people guessing how your firm works, because nobody asks for what they don't know exists. Give every contractor your firm's process on day one, keep client records locked, and they ramp up faster.
Everyone says data is the key to AI, and that the first step is connecting every system you have. For most firms that's the slow, expensive path. The faster win is writing down how the work runs and who decides, so people stop interrupting each other to find out.
AI assistants go wrong in ten well-known ways, and the security industry keeps the list. Here's the OWASP Top 10 for AI, explained for a firm without an IT department, and what we do about each risk.
Is one of these keeping you up at night?
Tell us what you're weighing. We'll say whether it's a problem we solve, and if it isn't, we'll say that too.