Frequently asked questions

Evoriqa is a platform for building AI customer-support agents trained on your own knowledge. They answer instantly across every channel, capture leads, book appointments, take actions in your tools, and hand off to a human when needed, all from one dashboard.

Paste your website URL so it crawls your content, brand the chat widget, then drop one script tag on your site. No engineers, no migration, and most teams are answering their first real question the same day.

No. Everything is point-and-click: training, branding, channels, and actions are all configured in the dashboard. The only 'code' is a single embed snippet you paste before your site's closing </body> tag.

Sign up at app.evoriqa.com and you're in — the Free plan needs no credit card and no sales call. Add your website or help docs as a knowledge source, try the agent in the Playground, then embed it and connect the channels you use. Every plan, including the white-label reseller tiers, can be started the same way.

Yes, twice over. Before you sign up, paste your website address on our Try page and we build a working agent from your own public pages, on a link you can share, with nothing to install. Once you have an account, the built-in Playground lets you chat with your agent privately, tune the prompt and knowledge, and see exactly which sources each answer came from before you ever go public.

Only from the knowledge you give it: your crawled website, uploaded documents, and FAQs. It uses hybrid retrieval, lets you review the sources behind each answer in your dashboard, and says 'I don't know' rather than inventing an answer when the information isn't in your knowledge base.

It's built to avoid that. Answers are grounded in your sources, and the agent is tuned to defer or hand off instead of guessing. You can review every answer's sources in the Playground's Debug mode and add missing knowledge in seconds.

Yes. Flip on the per-agent 'Show sources' toggle and the widget adds up to three source chips under each finished answer — links open in a new tab, uploaded files expand a short snippet. It's off by default, available on every plan, and the gate is server-side, so sources only appear once you've switched it on. Every chip is one of your own knowledge sources, never a link the agent made up.

Yes — turn on shadow mode (draft-for-review) per channel. The agent drafts every reply into your shared inbox, and nothing reaches the customer until a teammate reviews and sends it. It's the safe middle ground between human-first (the AI stays quiet) and fully autonomous, and the approval metrics — drafts, approvals, and how much your team edits each draft — show you when the bot has earned the switch to answering on its own. Available on WhatsApp, Messenger, Instagram, SMS, and Slack.

You measure them. With Live AI QA on (paid plans), an AI judge scores every production answer 0–100 shortly after it's sent — web, channels, and voice. Each reply carries its score in your inbox, a Low-QA filter lists the conversations with weak answers, and Analytics tracks the average over time. Vendors love to quote resolution rates; we'd rather hand you the per-answer numbers and let you see for yourself.

Yes. Every AI reply on a paid plan keeps a reasoning trace — expand “Why this answer” under the reply to see the exact knowledge passages it drew on with their similarity scores, the prompt it ran, and any actions it took. A bad answer stops being a mystery: you can see whether the wrong page was retrieved, the content is missing, or the question needed a human.

Yes. Agents can capture leads, book appointments, look up orders, create tickets, and call your own APIs through agentic actions, and every action is logged so you can see exactly what ran.

It hands off to a human with the full conversation context. Your agents get an AI copilot for suggested replies, plus canned responses, internal notes, SLAs, and CSAT, all in the shared inbox.

Yes. Agents understand and reply in the visitor's language, auto-detecting it from each message, so one agent can serve a global audience.

Completely. Set your colors, logo, position, greeting, and suggested questions with a live preview, so the widget matches your brand. On Pro and Business you can also remove the 'Powered by Evoriqa' badge.

Your website widget, WhatsApp, Messenger, Instagram, SMS, Slack, email, and voice (phone), every conversation flows into one unified inbox so nothing falls through the cracks.

Yes. The AI voice agent answers calls, speaks naturally with your callers, and drops a full transcript into the same inbox as your chats, so phone support scales without extra headcount.

Yes. Connect Cal.com, Calendly, or your own scheduling webhook, and the agent can check availability and book directly inside the conversation.

Through agentic actions and webhooks, your agent can call your CRM, order system, help desk, or any internal API. Leads and contacts are captured in Evoriqa and can be pushed onward automatically.

Webhooks are on every plan: point one endpoint per event group at your systems and Evoriqa sends signed, retried deliveries for leads, conversations, messages, and knowledge events, with a delivery log and one-click resend. A versioned REST API with scoped keys covers chatbots, knowledge, conversations, leads, channels, analytics, and the reseller surface; we switch API access on per account, so ask us and we enable it.

Point it at your website to crawl, upload documents (PDF, DOCX, TXT, MD, CSV), paste text, or add question-and-answer pairs. It indexes everything so the agent can retrieve the right passage for each question.

Schedule automatic re-crawls so answers track your latest content. Knowledge-gap detection surfaces the questions your agent couldn't answer well, so you know exactly what to add next.

It scales with your plan, from a single source on Free up to thousands on Business. Limits are counted active-only, so disabling a source frees the slot without deleting anything.

Yes. You set the agent's persona, tone, and instructions, choose how cautious it is, and decide which actions it can take. Because answers are grounded in your sources, it stays on-script for your business.

There's a Free plan, then Starter ($29/mo), Pro ($99/mo), and Business ($399/mo). Paying annually saves about 20%. See the pricing page for everything each plan includes.

Every paid plan guarantees a message floor on standard settings (Starter 2,400+, Pro 16,000+, Business 65,000+), backed by a monthly pool of AI credits (Starter 150k, Pro 1M, Business 4M) that covers chat, voice, and agent actions. A message on standard settings uses roughly 35–61 credits, an agentic action about 25, and a voice minute about 2,500 — voice also pays for telephony and speech, so Starter includes 60 voice minutes a month. Every rate is published on the pricing page.

Nothing is deleted and the agent keeps capturing conversations, handing them off to your team. Buy a one-time top-up pack (from $29 for 150,000 credits) — each pack is priced at the same per-credit rate as the plan of the same size, so your own plan's pack carries no overage penalty — or upgrade. Purchased pack credits never expire and carry over. Your plan allowance is a monthly reset, not a running balance — whatever you don't spend expires when the new month lands.

Yes. Set per-workspace spend caps, and for resellers, per-client caps, so usage and costs never surprise you. The Free plan runs on a fixed monthly message quota with no overage.

Absolutely. Upgrade, downgrade, or cancel from the Billing page at any time. Upgrades are immediate and prorated; downgrades take effect at your next renewal; if you cancel you keep access until the period ends, then move to Free. See our Cancellation & Refund policy for details.

Monthly plans are billed in advance and non-refundable for the current period, but you keep access until it ends. Credit top-up packs are non-refundable, but they never expire, so they're never wasted. Annual plans are billed in advance for the year and are non-refundable — run a month on a monthly plan first if you want to size your usage before committing. Either way the credits arrive one month at a time, on each monthly anniversary — an annual plan is paid up front but paced monthly, not granted all at once.

Yes. With white-label you put your brand, domain, and emails on the product, provision a workspace for each client, and present the whole experience as your own. Your clients see your brand, your domain, and your emails end to end.

Reseller plans give you a shared pool of AI credits and multiple client seats at agency pricing. You set your own prices for clients and keep the margin. Evoriqa handles the platform, you own the relationship.

Record what you bill each client and every client row shows this month's cost at your pooled-plan rate and your net margin live — with a short burn sparkline, red when a client is under water and amber under 20%. When a client's margin slips below 20% the org owner gets a heads-up email, re-armed whenever you update the billed amount; with no billed amount recorded the row shows '—' rather than a fake 100% margin. For the full story, each client has an on-demand ROI report — deflection, hours and cost saved, CSAT, leads, and appointments — exportable as CSV or an agency-branded PDF to hand over at renewal.

Client workspaces draw from your shared credit pool, and you can set per-client spend caps so no single client runs away with usage. You can see and manage every client from the reseller console. Running low? Buy a one-time top-up pack straight into your pool, or move to a larger tier — either way it takes effect immediately.

Yes. Every client workspace has its own knowledge, branding, channels, agents, and team, fully isolated from the others, while you manage them all from one place.

Yes, secrets are encrypted at rest (AES-GCM), 2FA is enforced, there's a full audit trail, and your data is kept under strict per-workspace tenant isolation. See the Trust Center for the full picture.

No. We never train foundation models on your content. Your knowledge and conversations are used only to power your own agents.

The tooling GDPR asks for: data export, right-to-erasure, per-workspace retention windows with automatic purging, and a DPA with a subprocessor list on request. Compliance also depends on how you configure and use the platform, so we're happy to walk your team through the controls.

Both, in the open. Our Trust Center sets out certifications with honest status (nothing is listed before it is attained), the security controls, our no-training-on-your-data commitment, retention and residency, and the sub-processor list by category (the named vendor list travels with a signed DPA) — and the Data Processing Addendum is published in full so your legal team can read it before signing. Every white-label brand also gets its own in-product trust page with a built-in Request-a-DPA form; on an agency's own domain that request routes straight to the agency, otherwise to us.

Our infrastructure isn't tied to a single region, but data residency is not guaranteed unless it's specifically agreed in writing. If you need a particular region, or a completed security questionnaire, email help@evoriqa.com and we'll work through it with you.

Yes. Invite teammates and assign roles (admin, member, viewer) so the right people can manage settings, work the inbox, or just observe. Seats scale with your plan.

Yes. The shared inbox lets agents claim and assign conversations, leave internal notes, use canned replies, and collect CSAT, with the AI copilot drafting replies alongside them.

Every plan has a Support section right in your dashboard: open a request with our team, keep the full threaded history, attach screenshots and files, and get notified by email and in-app the moment we reply. Business adds live chat with our team, on top of priority support and onboarding. Resellers support their own clients under their own brand, and we support your agency directly.

Email help@evoriqa.com for product help, or use the contact page to reach sales about reseller plans, security reviews, or anything else.