Questions, answered honestly.
The full objection library — cost, safety, process, ownership, and fit. If your question isn't here, a 15-minute intro call gets you a straight answer, no pitch.
The Foundations Session is A$300 at the founding rate (A$600–750 standard), credited toward your install. The build itself is a fixed price quoted in writing after scoping — before any work starts — with payment split 50% to commence, 50% on handover. Full structure on the pricing page.
Model usage plus any tool subscriptions — for most single-workflow agents roughly A$50–300 a month. We spec the expected running cost before the build so there are no surprises on your first invoice.
Sometimes you should — and we'll say so. The comparison to run, stated respectfully: an admin role costs a salary plus super, management, leave, and turnover; an agent is a one-time fixed build plus modest running costs, works 24/7, and is owned, not employed. The Foundations Session does this math with your actual numbers.
Every agent goes through a supervised proving run first: it drafts, a human approves, and every miss becomes a fix before go-live. In production it runs with logging, alerts, and a human escape hatch on anything outside its lane. Guardrails are part of the spec, not an afterthought — the full gate table is on The Cage page.
Agents run in your infrastructure with least-privilege access — only the permissions the workflow needs, with approval gates kept on for anything sensitive. Data residency is a deployment choice you own: your cloud region, your server, or your own hardware.
No. Model APIs are called under your accounts, with zero-data-retention options configured where the provider offers them. What each agent can read, and where data flows, is named in the spec.
That fear is why the proving run exists. Nothing client-facing sends without approval until that category of work has earned autonomy on your sign-off. Money, VIPs, and bulk actions never earn it — a human approves those every time.
Software that does a defined job the way a team member would: it reads the incoming work (an email, a lead, a ticket), decides what to do using your rules and your knowledge base, takes the action in your actual tools, and logs what it did. Not a chatbot widget — a worker on a workflow.
Typically 3–4 weeks for a single-workflow agent: scoped, built against your real historical cases, then run supervised until it's earned trust. Bigger scopes take longer, and the spec says so up front.
No — that's the point of the model. You describe the work in plain language; we translate it into a spec, build it, and hand it over running. Day-to-day, using the agent looks like reading messages in tools you already know.
A chat window waits for you to type. An agent logs into your systems, does the work, and reports back. If a chat window — or a $30/month SaaS tool — solves your problem, we'll tell you in the first call and we're done.
Yes. Everything runs in your accounts with your API keys, and the handover includes the code, credentials, runbook, and security config. If we part ways, it keeps running and you keep it all. No hostage data, no platform lock-in. The Method page shows the handover inventory.
Correct, and worth being precise about. You own the system: the code, the workflows, the integrations, your data, the security rules. The model underneath is a metered utility — and because you own the system, it's swappable: change providers or run a local model on your own hardware without rebuilding. What you can never be held hostage over is the part we hand you.
Everything is built to survive its builder: versioned repo, runbook, documented architecture, your infrastructure, your credentials. Compare that to an agency where "the team" is a rotating cast of juniors and the system lives in their accounts.
No, and we won't build it to. WeBuild agents take the repetitive work — the re-typing, the chasing, the after-hours phone — and hand the judgment back to your team. Operators use the recovered hours to grow, not to cut.
Rented flows move data between apps and break silently — and you can't take them with you. An agent owns an outcome end-to-end: it reads, decides, acts, and logs. And you own the agent.
We're new, and we don't fake it. Instead: the work so far is described honestly, stage by stage; the Founding Five program exists precisely because you're early; and you own everything, so your downside is capped at zero lock-in.
Still unsure?
Book a 15-minute intro call — no pitch, just a straight answer.
Book an intro call →30 minutes · no pitch · if AI isn't the answer, I'll say so