AI customer support that
refuses to lie.

Within a minute of checkout you will get an email with a secure link to your private upload page. Paste in your help center, FAQs, or SOPs, and we run an AI audit on the spot. We currently review every submission ourselves before activation, so you can also reply to that email if you have questions.

Typically within 24 hours of upload. Clean, well-structured docs can be live the same day. You can track progress in real time on a status page that we link from your email.

Yes. Color, logo, header copy, welcome message, theme, and starter questions are all customisable per tenant. Reply with your brand color (hex) and a logo URL after checkout and we'll match within 24h.

Send us the updated docs and we'll re-ingest your knowledge base, usually within a few hours. Self-serve doc management is on the roadmap once we have enough customers to design it right.

We use Google's Gemini API on the paid Tier 1 plan, which prohibits using your data to train any model. Your documentation and conversation logs are stored in Supabase (AWS eu-west-1, Ireland). Queries are processed by Google's Gemini API in the US under EU-US Standard Contractual Clauses and discarded after the response is returned. See our privacy policy for the full data flow.

A conversation starts when a visitor begins a new chat session with your widget. Follow-up messages inside that same session count as part of the same conversation, not as separate conversations.

Yes, for normal human support traffic. To protect your widget and our infrastructure from bot abuse or DDoS attacks, Pro includes a hard technical limit of 2,000 queries per day (about 60,000 per month). If your team genuinely exceeds this, we can raise the limit for higher-volume use cases.

Yes, in a weekend. Then spend the next 50 weekends on prompt regression tests, refusal calibration when the model updates, embeddings cost monitoring, widget hosting, CORS for customer domains, handoff routing, conversation logs schema, GDPR-compliant data retention, abuse rate limiting, prompt injection defense, and the dashboard still on your roadmap. We charge $99/mo because that's cheaper than your weekends.

Most answers return in 2–4 seconds.

Yes. Every plan includes a personalized analytics dashboard that ships with your widget snippet. You see total queries, resolution rate, average confidence, and recent handoffs, all scoped to your tenant. Full CSV exports remain available on request. Preview the Operator View →

Cancel anytime through Polar (our payment provider). No contracts, no cancellation fees. 7-day money-back guarantee on your first payment.

Prices are in USD and shown excluding VAT or sales tax. At checkout our payment provider, Polar, acts as merchant of record and adds any VAT or sales tax required for your location, so we handle global tax compliance for you. EU businesses with a valid VAT ID are reverse-charged: enter your VAT ID at checkout and no VAT is added. Either way, the net price you see here is what the subscription itself costs.

TrustQueue operates as a data processor under GDPR Article 28. Customer conversation data is stored in Supabase (AWS eu-west-1, Ireland). LLM inference runs on Google's Gemini API in the US under EU-US Standard Contractual Clauses; queries are not retained by Google after the response. We don't sell data, we don't use it for model training, and we honor deletion requests within 30 days. A Data Processing Agreement is available on request. See our privacy policy for the full subprocessor list and data flow.

Your knowledge base content stays yours. On request, we'll export your conversation logs and KB files in standard formats. We're bootstrapped and profitable by design. No VC runway risk.