FAQ
If yours isn't here, write us at the contact page or just book the audit — we'll answer it then.
General
Two to four weeks for a standard voice agent or chatbot. Custom workflows and integrations: 4–8 weeks. We don't ship until you sign off on a transcript review.
Every project starts with a Free AI Audit — 30 minutes with the founder where we map three places AI can save you money. If a pilot makes sense, we ship the smallest useful thing in 2–4 weeks and tune it weekly. You only continue if it's working.
Honestly — for the most repetitive 30–50% of their work, yes. The intent is to free your team for higher-value work, not to delete jobs. Most clients see headcount stay flat while output goes up.
Encrypted at rest and in transit. Subprocessors documented. We never train models on your data. HIPAA BAAs available for healthcare. Audit logs on every PII access.
Every agent we deploy offers a human at any point. We track 'human request rate' as a quality metric. If it's high, the bot is the problem — not your customers.
ChatGPT can't see your CRM, your calendar, or your inventory. Our agents are grounded in your real systems and configured for your specific workflows. ChatGPT is a tool. We build systems.
Working with us
Yes — fully. GitHub repo transferred at launch for web work, and you own the prompts, evals, and infrastructure-as-code for AI builds. No license fees, no platform lock.
We start every engagement with a paid two-week prototype or audit so you see real output before committing to production. If the prototype doesn't clear our quality bar — or yours — we'll say so. You only pay for what you got.
Yes. We hand over documented runbooks, code, evals, and credentials. Many engagements end with your team running the system; we stay on for advisory or new work.
No. We never train models on your data, and we never grant model providers training rights to it.
Every agent we deploy has audit logs, decision rationale logging where applicable, and configurable retention. We design with explainability for the cases that need it.
Service-specific
A representative question per practice. Each service page has the full list.
Yes — we never deploy an agent that pretends to be a human. It introduces itself as your automated assistant and offers a human at any point.
It's grounded in your actual content and connected to your real tools. ChatGPT-the-product can't see your inventory, your CRM, or your calendar. Ours can.
Usually no. We orchestrate the tools you already pay for. If a tool is the bottleneck we'll say so and recommend a swap — with reasoning, not a kickback.
Not for most SMBs. We can run on managed Postgres or BigQuery free tier. If your data outgrows that, we'll tell you.
No — and they shouldn't. The voice screen identifies itself, is short, and offers a human at any point. We measure candidate experience as a KPI.
Yes — fully. GitHub repo transferred at launch. No license fees, no platform lock.
Then we tell you, and you don't spend production money on it. The prototype fee covers an honest read — including 'this isn't worth building.'
Free AI Audit
We'll show you three places AI can save you money this quarter — yours to keep, no obligation, no slide deck.
