1.1
AI Chatbots
Agents that understand context, learn from your knowledge base, and answer accurately around the clock in 11 Indian languages.
AI chat & voice agents · live in minutes, not months
Customer conversations
that resolve themselves
Qvrone deploys chat and voice agents that read your real order, booking and customer data — then answer, act, and hand off to a human with the full transcript when they should.
Chat agentsVoice agentsShared inboxKnowledge baseAutomation flows
hey, where's my order? placed it tuesday
tool callshopify.orders.lookup({ email, since: "-7d" })Found it — order #4821 shipped Wednesday evening via BlueDart. It's out for delivery today and should reach you by 6 PM.
can i change the delivery address?
Not once it's out for delivery. I can reschedule to tomorrow at a new address instead — want me to do that?
→ intent: order_status → tool: shopify ✓ → answered → escalation: none · 840ms
- 11languages
- Indic + English, one agent
- 6channels
- Web, WhatsApp, IG, Slack, email
- 50+templates
- Pre-built conversation flows
- <1sfirst token
- Streamed from your knowledge base
Fig.1 · The platformOne platform behind
One platform behind
every customer conversation
Qvrone combines chat and voice agents with a shared inbox, knowledge base and automation flows — so support, sales and operations run on the same context instead of three disconnected tools.
1.2
Voice Agents
Human-like voices that handle inbound and outbound calls, schedule appointments, and qualify leads without a script that shows.
1.3
Multilingual
Hindi, Tamil, Telugu, Bengali, Marathi, Gujarati, Kannada, Malayalam, Punjabi and more — one agent, no separate setup per language.
1.4
Two-minute setup
No code. Pick a template, point it at your data, publish. Then A/B test wording and routing without redeploying anything.
1.5
Enterprise security
SOC 2 aligned controls, encryption in transit and at rest, GDPR-ready data handling, and data residency in India.
1.6
Always on
Queries never queue overnight. Agents run across time zones and hand off to your team the moment a human is genuinely needed.
Qvrone is in limited release — panel access is granted per account rather than on signup.
Fig.2 · ServicesStart from a template.
Start from a template.
Ship the same afternoon.
Fifty-plus conversation templates, each already wired to the systems it needs. Pick one, point it at your data, and it is answering customers before the day is out.
2.1
Appointment Booking
Reads live availability, books the slot, sends the reminder, and handles the reschedule when it comes.
Google CalendarSMS
Request this template2.2
FAQ Responder
Answers hours, services, pricing and policy questions from your knowledge base — cited, not improvised.
Knowledge baseWeb widget
Request this template2.3
Lead Qualification
Asks the qualifying questions, captures contact details, scores the lead, and routes hot ones to your team.
CRMSlack
Request this template2.4
Order Status Tracker
Pulls the real order record and returns carrier, dispatch time and delivery window — never a canned reply.
ShopifyWooCommerce
Request this template2.5
Payment Reminder
Places outbound voice reminders for overdue invoices, records the outcome, and retries on your schedule.
VoiceInvoicing
Request this template2.6
Health Intake
Collects symptoms, history and insurance before the appointment, so the clinician opens a complete chart.
EHRHIPAA mode
Request this template2.7 — Industry workspaces
Or take the whole vertical, pre-configured
A workspace bundles the templates, integrations, escalation rules and compliance defaults for one industry — so you are tuning, not assembling.
Fig.3 · ProcessFour steps between here
Four steps between here
and answering customers
No integration project, no professional services engagement. The longest part is deciding which questions you want handled first.
Step 01: Connect your data
Point Qvrone at the systems that already hold the answers — your storefront, CRM, calendar, help centre, or a folder of PDFs. Indexing starts immediately; nothing has to be rewritten.
shopify · hubspot · google-calendar · zendesk · url · pdfStep 02: Pick an agent
Choose a template or an industry workspace. It arrives with the prompts, tool permissions and escalation rules for that job already set, so you are editing wording rather than building from nothing.
template: order-status → tools: [orders.lookup, shipping.track]Step 03: Set the guardrails
Define what the agent may do on its own and where a human signs off. Refunds over a threshold, anything touching medical records, any action you name — those pause for approval instead of proceeding.
risk: high → approval_required: true → notify: #supportStep 04: Go live and tune
Publish to every channel at once. Watch real transcripts, see which questions escalate, and correct the knowledge base in place — the agent picks up the change without a redeploy.
web · whatsapp · instagram · slack · email · voice
Fig.4 · In motionEvery question answered,
Every question answered,
while you sleep
Your agent handles the repetitive eighty percent — order status, hours, bookings, policies — and hands the rest to your team with the full context attached.
Support Agent
online · replies instantly
Answers in under a second
4.1Streaming responses from your own knowledge base. No 'let me check on that' — the answer is already there.
Speaks your customer's language
4.2Multilingual out of the box, including Indic languages. One agent, every customer, no separate setup.
Connected to your systems
4.3Reads real order status, bookings and customer records from Shopify, your CRM and your calendar — not canned replies.
Hands off when it should
4.4Recognises when a conversation needs a human and escalates with the full transcript attached. No repeating themselves.
One agent, deployed everywhere your customers already are
Website widgetWhatsAppInstagram DMSlackEmailShopify storefront
Fig.5 · Connected & governedIt plugs into the stack
It plugs into the stack
you already run
An agent is only as useful as the systems it can read. Qvrone ships native connectors for the tools your business already depends on — and a governance layer over everything it does with them.
5.1 — Native connectors
Commerce
- Shopify
- WooCommerce
- BigCommerce
- PrestaShop
- WordPress
CRM
- HubSpot
- Salesforce
- Pipedrive
- Zoho
Messaging
- Slack
Marketing
- Klaviyo
- Mailchimp
- ActiveCampaign
- Omnisend
- Brevo
- MailerLite
Support & automation
- Zendesk
- Judge.me
- Zapier
Something missing? Any HTTP API becomes a tool your agent can call, with the same permission and audit rules as a native connector.
5.2 — Security & governance
Encrypted end to end
TLS in transit, AES-256 at rest. Conversation payloads and knowledge-base documents are encrypted on the same terms.
Data residency in India
Customer data can be pinned to Indian regions, so it never leaves the jurisdiction your compliance team answers to.
Human approval gates
Name the actions an agent may not take alone — refunds above a threshold, anything touching records — and they pause for sign-off.
Every action audited
Each tool call, escalation and approval is logged with actor, timestamp and payload. Exportable when an auditor asks.
Scoped credentials
Connectors hold least-privilege tokens per workspace. An agent can read the order it was asked about and nothing beyond it.
PII redaction
Card numbers, government IDs and contact details are masked before a transcript reaches logs, analytics or model context.
Fig.6 · BillingYou pay for what
You pay for what
the agents actually use
Two independent meters, no seats and no tiers. Rates are still being set, so we are not publishing a number we would have to change — talk to us and we will quote against your actual volume.
6.1
API
Metered on model usage
Call the agents from your own product. You are billed for the tokens the routed model actually consumed on each turn — input and output, at that model's rate. Cheaper models handle classification and routing, so the cost of a turn tracks its difficulty rather than a flat per-conversation fee.
- Input and output tokens, per model
- Billed per turn, not per conversation
- Routing to a cheaper model costs less
- Tool calls and retrieval included
- No minimum, no seat count
6.2
Cloud
Metered on workspace resources
Run agents on our infrastructure instead of your own. You are billed for what the workspace holds open — runtime hours, the vector store behind your knowledge base, transcript retention and egress. An idle workspace draws nothing.
- Agent runtime, per compute-hour
- Vector store, per GB-month
- Transcript retention, per GB-month
- Egress, per GB transferred
- Idle workspaces are not billed
Rate card
Per-model and per-resource rates are being finalised. Send us the volume you expect — turns per month, retention window, whether you need the API, the cloud, or both — and we will come back with real numbers rather than a range.
Fig.7 · QuestionsThe things people ask
The things people ask
before they sign up
Answer questions from your knowledge base, and take actions through connected tools — look up an order, book or move an appointment, check a booking, update a record, place an outbound call. You decide which of those it may do unattended and which pause for a human.