FAQ
The honest answers to what people ask before they deploy RAD — how the human gate works, where your data lives, what it costs, and what's specific to each offering. Short, candid, no fine print.
RAD solutions comes in two offerings, both human-gated and both deployed onto your own cloud. RAD Business runs everything a business does to find, win and keep customers — content, leads, campaigns, a website chatbot, surveys and CRM sync. RAD Professional gives each executive a private chief-of-staff over their own Google Workspace — inbox, calendar, briefings, drafting and research. The questions below start with what's true for both, then cover what's specific to each.
No. Every action that has an effect in the world — an email, a social post, a campaign, a published article, a booked meeting, a shared document, a new task — is staged as a draft and waits for a human to approve it. The part of the system that writes is deliberately separate from the part that acts, with your approval in between. If nobody approves a draft, nothing happens. There is no "autopilot" mode to leave on by accident — that isn't a setting you trust, it's how the system is built. Read-only things, like summarising an inbox or preparing a briefing, just happen, because they change nothing.
No. The AI reasoning runs through governed API gateways — Anthropic's Claude via a LiteLLM gateway (RAD Business also uses Google's Gemini for retrieval), and that API traffic is not used to train the models. That's contractual, not a checkbox. We're happy to send our security write-up so your IT person or partner can read the specifics.
Yours. The whole system deploys onto your own cloud account — for RAD Business, your Google Cloud project; for RAD Professional, inside your own Google Workspace through your own login. Your customer list, leads, mail, calendar, files and content stay in your cloud, under your control. No third-party SaaS holds your data. The AI reasoning calls out to Claude and Gemini through governed APIs, but the data itself never leaves your account.
It drafts from your own material and how you actually write, not a generic template, and your team corrects it during an initial period — "too formal," "shorter," "don't say that." It adapts, and within a couple of weeks most drafts need little editing. Because a human reviews everything, an off-voice draft is one you catch before it ever ships.
We scope it to your engagement, but the pattern is the same: a setup phase to deploy it onto your cloud and connect it with your permission, then a first stretch where your team corrects the drafts freely so it learns your voice. We start narrow — a couple of capabilities — and switch on more as you build trust. You don't have to turn everything on at once.
It's scoped per engagement, because it depends on which capabilities you want and the size of your setup. We'd rather understand your needs first and come back with a real number than quote something generic. What we can say is there's no per-seat SaaS markup on your own data — it runs on your own cloud, so you're not renting your data back from a vendor every month.
No. We — or your agency/IT partner — handle the deployment and connect it to your cloud with your permission. Your team only ever touches the simple approval surface and works with the drafts.
You read every outward action before it happens, so a wrong draft is one you correct or reject before it ever reaches anyone. Nothing sends until a person approves it.
Yes. It runs on your own cloud and connects through your own login, so you stay in control. You can switch off any capability, revoke access, or remove it entirely — and your data was never anywhere but your own cloud to begin with.
A human-gated AI system that runs your customer-facing marketing and communications from one engine: blog and newsletter content, social, lead capture and nurture, inbound-inquiry replies, bulk-email campaigns, a website chatbot, surveys, market intelligence, and CRM sync. The AI drafts everything; a person on your team approves it; then it sends or publishes. Nothing goes out on its own.
Draft and distribute blog posts, newsletters and social content; capture leads from your site and chatbot and run timed follow-ups; read inbound inquiries and draft replies; run human-approved bulk-email campaigns; answer visitor questions on your website from your own content with citations; collect survey feedback; scan the web for buying-intent and competitor signals; and keep your CRM in sync — all gated by human approval before anything customer-facing goes out.
In a simple review screen (your Directus data layer). You open the relevant queue — leads, content, campaigns — read or edit the AI's draft, and set it to Approved (or Rejected). A separate scheduled sender then dispatches only what's approved. Reviewing drafts is the one routine manual step.
Instead of a blog platform plus an email tool plus a social scheduler plus CRM glue plus assorted "AI marketing" add-ons, you run a single system that does all of it, on infrastructure you own, with one approval queue and one brand voice. Fewer subscriptions, fewer logins, one source of truth.
The chatbot answers from your own approved content, with citations, and it can't email or post. When it captures a lead from a conversation, that lead still flows into the same approval queue — so any outreach it triggers is a draft a human reviews, not an automatic message.
No — auto-posting to third-party platforms violates their terms, so we don't. The system prepares the post and a person publishes it. For your own channels (your blog and newsletter), it publishes drafts after approval.
It's designed to consolidate the stack — drafting, follow-up, campaigns, chatbot and CRM sync in one engine on infrastructure you own. Some clients replace most of their tools; some keep a favourite or two. If you work with an agency, it usually makes them faster — many of our partners are agencies who run this for their clients.
A private digital chief-of-staff for one executive, working inside your own Google Workspace. It triages your inbox, runs your calendar, briefs you before meetings, drafts in your voice, answers questions from your own files, and runs research — and it never sends, books, or commits anything in your name without your approval.
Inbox triage and reply drafting; calendar agendas, scheduling and reschedule proposals; morning and pre-meeting briefings; drafting emails, memos and documents; answering questions from your own Drive and Docs with citations; and private web research. Tasks and document production round it out.
Conversationally, in a chat thread — you talk to the assistant in Telegram, Slack or Google Chat, and you approve or redirect in plain language: "approve," or "change the time to 3pm." Every proposal and decision is also recorded in an immutable ledger, so there's a complete audit trail.
A chat app — Telegram or Slack natively, with Google Chat available via a bridge. You message it the way you'd message a colleague.
Three things. It's Workspace-native — it acts on your real inbox, calendar and Drive, not a chat window you copy-paste into. It's gated — it acts as you, so it never sends or books without your explicit approval, which a general chatbot isn't built to guarantee. And it's sovereign — it runs on your own cloud with your own login, so your mail and files don't leave your control. General chatbots are brilliant thinking tools; this operates your day under your authority.
No — it augments judgement, it doesn't own decisions. It handles the draft-everything, prep-everything, never-forget layer; the decisions and the human relationships stay with you and your EA. If you have a great EA, this gives both of you leverage. It has no authority to decide whether to act — only to propose.
No. It's granted least-privilege access, function by function, and the knowledge base ingests only the folders and documents you choose — not your whole Drive wholesale.
Yes — it's built natively for Google Workspace (Gmail, Calendar, Drive, Docs, Tasks). It isn't a fit for Microsoft 365 today. If that changes, we'll reach out.
Still have a question?
Whether you're weighing RAD Business for your organisation or RAD Professional for your leaders, the next step is a focused, no-obligation conversation. We'll answer the specifics and, if it's a fit, scope a narrow first engagement.
Takes a minute — we'll email you back to confirm a time.