How to build a 7x24-hour AI customer service robot with OpenClaw (zero code)

Let's be honest — most "zero code" promises in tech are half-truths. You still end up wrestling with YAML files, environment variables, or some obscure config syntax. But OpenClaw on Tencent Cloud Lighthouse is genuinely one of the rare cases where you can go from nothing to a working 24/7 AI customer service bot without writing a single line of code.

I'm not exaggerating. The entire setup happens through a visual console panel and an interactive CLI wizard that asks you questions in plain English. If you can click buttons and paste an API key, you can do this.

Here's the full walkthrough.

What You're Building

By the end of this guide, you'll have:

An AI-powered customer service agent running on a cloud server
Connected to one or more messaging platforms (WhatsApp, Telegram, Discord, etc.)
Responding to customer queries 24 hours a day, 7 days a week
Completely isolated from your personal computer and data

The agent uses a large language model (like DeepSeek, GPT, or Claude) as its brain, and OpenClaw as the orchestration layer that manages conversations, memory, and channel integrations.

Why Cloud, Not Local?

I get asked this a lot. "Can't I just run it on my MacBook?"

Technically, yes. Practically, please don't — at least not for production customer service. Here's why:

Uptime: Your laptop sleeps, updates, and travels. A cloud server doesn't.
Security: OpenClaw has system-level access. On your personal machine, that means your photos, documents, and browser sessions are all exposed.
Cost: A Lighthouse instance costs less per month than the electricity to keep a Mac Mini running 24/7.

Step 1: Launch Your Instance (2 Minutes)

Go to the Tencent Cloud Lighthouse Special Offer page:

Visit the page to browse pre-configured OpenClaw instances.
Choose the "OpenClaw (Clawdbot)" template under the AI Agent category.
Deploy by clicking "Buy Now" to spin up your server.

Select at least a 2-core, 4GB RAM plan. For international customer service (WhatsApp, Telegram), choose an overseas region like Singapore. For domestic Chinese platforms, pick a mainland region.

That's it for infrastructure. Your server comes with OpenClaw pre-installed.

Step 2: Configure Your Model (1 Minute)

In the Tencent Cloud Console, navigate to your Lighthouse instance and click Application Management. You'll see a clean visual panel.

Under Models, paste your API key for your chosen LLM provider. Options include Tencent Hunyuan, DeepSeek, OpenAI GPT, Google Gemini, Anthropic Claude, and more. Full details on supported models are in the Custom Model Tutorial.

Click Add and Apply. Wait for the status to show "in use." Done.

Important: Never share or hardcode your API keys. The Lighthouse console stores them securely on your instance.

Step 3: Set Up Your Channel (2 Minutes)

Now connect OpenClaw to your customer-facing messaging platform. Open the server's terminal via OrcaTerm (Tencent Cloud's built-in web terminal — no SSH client needed).

# Start the interactive configuration wizard
clawdbot onboard

The wizard walks you through everything with arrow-key navigation:

Disclaimer → Yes
Mode → QuickStart
Config → Use existing values
Model → Skip (already configured in console)
Channel → Pick your platform (WhatsApp, Telegram, Discord, etc.)
Skills → No (keep it simple for now)
Hooks → Select only session-memory
Restart → Yes

For platform-specific walkthroughs:

Step 4: Pair Your Bot (30 Seconds)

After restarting, go to your messaging app and send a message to your bot. You'll receive a pairing code. Back in the terminal:

# Replace <code> with the actual pairing code you received
# Replace <channel> with: telegram, discord, or whatsapp
openclaw pairing approve <channel> <code>

Pairing confirmed. Your bot is live.

Step 5: Enable 24/7 Background Mode

The final piece — make sure the agent keeps running after you close the terminal:

# Enable persistent services for your user
loginctl enable-linger $(whoami) && export XDG_RUNTIME_DIR=/run/user/$(id -u)

# Install and start the background daemon
clawdbot daemon install
clawdbot daemon start

# Confirm it's running
clawdbot daemon status

If the status is healthy, you're done. Close the browser, shut your laptop, go to bed. Your AI customer service agent is now handling inquiries around the clock.

Zero Code — But Not Zero Brain

Just because you didn't write code doesn't mean you can't customize the agent's behavior. Through the configuration, you can:

Set a system prompt that defines the agent's personality, knowledge base, and escalation rules
Choose which model powers the responses (balance cost vs. quality)
Enable skills later for advanced capabilities like web browsing, email, or file handling
Connect multiple channels simultaneously from a single instance

For skill installation, check out the Skills Guide.

Your Move

Building a 7x24 AI customer service robot used to require a dev team, infrastructure expertise, and serious budget. Now it takes one cloud instance, one API key, and five minutes of clicking through a wizard.

Get started at the Tencent Cloud Lighthouse Special Offer:

Visit the page for dedicated OpenClaw pricing.
Choose the OpenClaw (Clawdbot) app template under AI Agent.
Deploy via "Buy Now" and have your agent live before your coffee gets cold.

No code. No excuses. Just a tireless AI agent working for your customers while you focus on everything else.