Every question
we get asked.

Most agents are live in 72 hours after your demo call. Complex integrations (custom CRMs, multi-location calendar logic, regulated workflows) can take up to a week — we tell you up front if your build is in that bucket.

A 20-minute demo call. You see a live agent in the first few minutes, then we walk through the questions we need to build yours: services, offers, pricing, tone of voice, FAQs, escalation rules, and which channels you want. The call is the actual on-ramp — there's no separate web form to fill out before or after.

No. The 20-minute demo call is the intake — we cover the questions live so there's no long form on your side. If something comes up after the call, we email you a quick question and you reply when convenient.

No. You need a business — services, offers, FAQs, a channel customers contact you on. The web chat widget is one of three channels and is optional; many customers run only Instagram or Messenger.

No. The Saygents team handles the build, training, and channel wiring. The dashboard is designed for business owners, not engineers. The most technical thing you'll do is paste a one-line snippet onto your site if you want the web chat widget — and we send you that pre-filled.

Humans on the Saygents team using AI tooling. Not a no-code wizard you fill out. Not a drag-and-drop builder. We do the work; you get a live agent.

On the demo call we walk through your services, offers, pricing, tone, FAQs, and any documents you want to share (PDFs, brochures, internal SOPs). The agent is trained on that material — it does not draw answers from a generic model.

Yes — the operational settings are all yours from the dashboard: agent on/off, business hours, away message, escalation keywords, blocked users, retention, and you can upload knowledge documents (PDFs, SOPs, brochures) any time. Most changes — updating services, prices, hours, tone, FAQs, escalation rules — ship within 12 hours of request. If you need a rush (urgent customer-facing fix), flag it as urgent and we'll prioritize.

File a modification request from the inbox — paste the original reply, suggest the better one, and add a note if it helps. The Saygents team reviews every request and applies approved changes. Standard turnaround is ~12 hours; mark a request urgent for a faster fix.

Three: Web Chat, Instagram DM, and Facebook Messenger. One agent runs one channel, but the same knowledge base, tone, and calendar travel across them.

One agent. Same brain, same knowledge, same tone, same escalation rules — running on every channel you turn on. A customer who DMs you on Instagram and then texts you the next day gets a continuous experience, not a hand-off between two different bots.

You connect from your agent's Connections page in two clicks — sign in with your existing Facebook account, pick your Page (Messenger) or Instagram Business account, grant permissions, done. No Meta developer account, no Business Verification. We handle the technical wiring on our side.

Yes. Saygents connects directly to Google Calendar, Microsoft 365 / Outlook, Cal.com, or Acuity. The agent reads real availability (respecting your hours, buffers, and team-member calendars on multi-rep setups) and writes real calendar events during the conversation. Calendly works as a share-link drop-in.

We push every captured lead to Slack and Zapier. Zapier covers thousands of CRMs and tools — set up a Zap once and every Saygents lead lands in Salesforce or wherever else you work. For teams that want their own routing logic, use Zapier's filters and branches. If you need a native integration (no Zapier in the loop), email info@saygents.com and we'll scope it.

Often, yes. If your system has an API or accepts webhooks, we can usually wire it in during onboarding or as an addition after launch. Heavy custom builds may extend the 72-hour timeline — we'll tell you up front.

English out of the box. The underlying model supports 30+ additional languages and can reply in whichever language the customer writes in — we tune the agent for the mix your business actually sees.

Three plans. Solo: $150/month for 1 agent on 1 channel with 500 conversations included. Duo: $250/month flat for exactly 2 agents — two channels of your choice — with 1,000 conversations pooled across the tenant ($125/agent). Trio: $300/month flat for exactly 3 agents — one for each channel (Web Chat, Instagram, Messenger) — with 1,500 conversations pooled across the tenant ($100/agent). All three plans are month-to-month, and the overage rate ($50 per extra 250 conversations) is the same on every tier. There is no add-an-extra-agent path — to cover more channels you move up a tier.

A conversation is a back-and-forth thread between one customer and your agent on a single channel within a single session. Reopening the same thread later in the day counts as one conversation; a brand-new thread is a new one.

Agents keep responding — there is no hard cap on conversations. Past your included pool (500 on Solo, 1,000 pooled on Duo, 1,500 pooled on Trio), overage bills at $50 per 250 extra conversations, rounded up, on the next invoice. All three channels — Web Chat, Instagram, and Messenger — count toward the same allowance and the same overage rate. We surface live usage in the dashboard so there are no surprises.

The first month is charged in full at signup and is non-refundable. That's because the 72-hour custom build happens before your agent ever serves a conversation — we hand-build it on day one. Subsequent months bill on the same day-of-month anchor.

Yes. Plans are month-to-month with no long-term contracts — cancel any month from the dashboard and your next renewal stops. Your agent stays live through the end of the month you have already paid for.

No. Because the custom build happens on day one — before the agent has answered a single conversation — we don't offer refunds on any month, including the first. We mitigate this by being aggressive about pricing transparency, by running the 20-min demo before you commit (so you see a live agent work first), and by making cancellation truly month-to-month.

We retry the charge once at +24 hours. If it still fails, the subscription is canceled and your agent stops serving conversations until you update billing.

Not currently. Saygents is monthly only — we'd rather earn the renewal each month than incentivize a long lock-in.

If you need a deployment larger than the Trio plan or a custom integration, email info@saygents.com and we will scope it.

Your customer data is yours — lead names, emails, phone numbers, conversation content, and any documents you upload all belong to you, not us. It is stored on infrastructure in the United States, each tenant's data is isolated from every other tenant's, and it is not used to train any AI model. You can request export or deletion of your data anytime. Full overview at /security and /data-deletion.

No. Your conversations are used only to run your agent — never to train any AI model, ours or our model provider's. We use Anthropic's commercial API, which by default does not train on data sent through it. Your data powers your agent and nothing else.

24 months by default. If you need a shorter retention window for compliance reasons, email security@saygents.com — we will scope it case by case.

United States. The full sub-processor list is at /security. If your jurisdiction requires data to remain in a specific region, email security@saygents.com before signing.

Saygents is built for general sales and support conversations. We are not HIPAA-compliant, we don't sign BAAs, and we don't offer a HIPAA-specific configuration — so the agent shouldn't be used to collect or handle protected health information, financial account details, or other regulated data. What we do offer: during onboarding we set explicit guardrails so the agent steers regulated topics to a human and avoids capturing sensitive information. If your business sits in a regulated space, email info@saygents.com first and we'll be straight about what the agent can and can't safely do.

Yes. Self-service deletion paths are documented at /data-deletion, and on request we will confirm deletion in writing.

Yes — and it's built so the agent doesn't improvise. It answers from the materials you give us, and escalates to your team whenever a question falls outside that, touches a sensitive topic (payment, complaint, refund, legal or medical), or the customer asks for a person. Escalation rules are configured per agent during onboarding; from the dashboard you adjust the escalation keyword list and business hours.

Yes. On the demo call we capture sample copy, common phrasing, and what to avoid. The agent is tuned to your tone, not the model's default. If anything sounds off after launch, flag the conversation in your Inbox and we'll retune within 24 hours.

Yes. Lead capture is one of the standard use cases. The agent can ask qualifying questions in sequence, branch on answers, and act on the outcome (book, escalate, or capture the lead for follow-up via Slack and Zapier).

Images: yes — the agent can interpret photos customers send (a damaged roof, a product they're asking about). Voice notes and files: in development; ask during onboarding for current support.

Yes. You can configure automated follow-up rules — for example, ping a lead 24h after they went silent on a quote conversation. All follow-ups respect channel rules (e.g. Meta's 24-hour messaging window on Instagram and Messenger).

Yes. Multi-location and multi-rep setups are supported on Trio (and beyond, on request) — each location/team member can have its own agent or share an agent with location-specific context. Calendar routing respects the right person.

Per-agent and per-channel: conversation volume, captured-lead count, bookings, escalation rate, average response time, top customer questions. Pulled into the dashboard; weekly digest email if you want it.

A combination: explicit topic guardrails defined during onboarding (the agent will refuse or escalate on listed topics), confidence thresholds for handoff, system-prompt-level rules about what the agent will never claim, and human review of conversation logs during the first week of every build.

Email info@saygents.com for general questions, security@saygents.com for procurement and security questions. Duo and Trio customers get a dedicated channel during onboarding for ongoing changes.

We target same-business-day response on support emails.

We target 99.9% on the agent serving path. Status is monitored continuously; incident reports are sent to affected tenants. We don't currently publish a public status page — on the roadmap.

Yes. Upgrades take effect immediately and prorate; downgrades take effect at the next renewal so you get the value of the month you've already paid for.

No formal "pause" today — the cleanest path is to cancel and re-onboard later. Re-onboarding from a previous tenant is faster than a fresh build because we keep your knowledge base on file for 90 days post-cancellation (unless you request earlier deletion).