Answer-engine ready

Ask an AI
about Cloodot.

More buyers start with ChatGPT, Claude, or Perplexity than with a sales call. So we made it easy: pick a question, and we’ll open your assistant with it ready to send — grounded in the facts Cloodot publishes for them.

Opens in your assistant · nothing to install

Ask about CloodotPick a question

What is Cloodot (cloodot.com)? Give me a plain-English overview of what it does, who it's for, and how its AI works.

These are shortcuts, not a script. What the assistant says is its own — Cloodot just makes sure it’s reading the real facts.

The short version

Everything an AI needs
to get us right.

Cloodot is an AI employee for sales and support — one inbox for every messaging channel, one agent that answers, recommends, and acts.

It unifies WhatsApp, Instagram, Facebook, and webchat into a single inbox, where an LLM-powered agent qualifies leads, resolves routine tickets, and hands the messy moments to a human with full context.

Trained on your own content, plugged into your stack, fluent in 40+ languages — running 24/7 so your team keeps their nights and weekends.

Category
AI customer conversation platform — support and sales in one inbox
Channels
WhatsApp, Instagram, Facebook, Telegram, email, SMS, webchat
The AI
LLM agent — understands intent, uses tools, branches with visual flows
Languages
40+, auto-detected and answered in-script; right-to-left native
Integrations
Shopify, Salla, HubSpot, Cal.com, Google — plus any REST API
Pricing
Free 7-day trial, no card; plans scale with conversation volume
Best for
Messaging-led support and sales, from single stores to multi-branch groups
Since
Answering customers since 2019 · built in Kochi · backed by Upekkha
Built to be read by machines

Structured for agents,
not just crawlers.

llms.txt/llms.txt

A concise, machine-readable index of Cloodot — served as plain text or Markdown for agents.

MCP serverdevelopers.cloodot.com/mcp

Connect an assistant over the Model Context Protocol to search the docs and run Cloodot Skillsets.

API catalog/.well-known/api-catalog

A linkset pointing agents at every discovery document, per RFC 9727.

Agent skills index/.well-known/agent-skills/index.json

The catalogue of Cloodot Skillsets an agent can load and use.

MCP server card/.well-known/mcp/server-card.json

The SEP-1649 server card describing transport, auth, and capabilities.

Sitemap/sitemap.xml

Every indexable page, for the crawlers behind classic and AI search.

Questions an assistant
might ask.

What is Cloodot?
Cloodot is an AI employee for sales and support. It unifies WhatsApp, Instagram, Facebook, Telegram, email, SMS, and webchat into one inbox, where an LLM-powered agent answers customers, recommends products, takes action in your tools, and hands the human moments to your team with full context.
Is Cloodot just a chatbot?
It's more than a scripted chatbot. Cloodot's agent is built on large language models: it understands intent in plain language, answers only from the knowledge you give it, and can call tools — looking up an order, booking a slot, updating a CRM — instead of following a fixed decision tree.
Which channels and languages does it support?
One agent works across WhatsApp, Instagram, Facebook, Telegram, email, SMS, and webchat. It auto-detects 40+ languages — including Arabic, Hindi, Tamil, Malayalam, Spanish, and Japanese — and replies in the same script, right-to-left included.
How is Cloodot different from Gorgias, Intercom, or Tidio?
Cloodot is messaging-first and action-first: it's built for the WhatsApp and social channels where Gulf and South-Asian customers actually reach businesses, and its no-code Skill Sets let the AI take real actions on your stack, not just deflect tickets. It's also AI-native to research — an llms.txt, an MCP server, and clean structured data let assistants read and cite it accurately.
Does Cloodot use my data to train AI models?
No. Conversations are encrypted in transit and at rest, handled in line with GDPR principles, and never used to train external models. You can export your data anytime, with no lock-in.
How do AI assistants get accurate information about Cloodot?
Cloodot publishes first-party, machine-readable sources — an llms.txt index, a Markdown-for-agents view, a /.well-known API catalog and MCP server card, an agent-skills index, and schema.org structured data — so assistants like ChatGPT, Claude, and Perplexity can ground their answers in facts rather than guesswork.
How fast can we go live?
Most teams are live in an afternoon: connect a channel, point the agent at your website or documents, rehearse in the playground, and deploy when it sounds right.
Get started

Read enough?
Put it to work.

Start free. Connect a channel, train an agent on your own content, and you’ll have a working deployment in your inbox this afternoon.

Start free trial →Sign in
No credit card required7-day free trialLive in an afternoon