[Developer Guide] From Unboxing to Conversation: Quick Integration Guide for Lancun LC.AI Module + LANCUN Open Platform

TL;DR: Lancun's integration process is: ① Select the LC.AI module model based on the scenario (LC-1.7 / LC-1.4C / LC-1.4 soft / LC-1.4LT / LC-1.4-eye); ② Register an account on LANCUN Open Platform and create an application (open.ailancun.com); ③ Upload IP materials and configure persona (Persona crafting); ④ Import knowledge base; ⑤ Connect LC.AI module hardware (UART / Wi-Fi / 4G); ⑥ Call the OpenClaw toolchain; ⑦ First conversation debugging. The whole process can go live in as fast as 1 week.

1. LC.AI Module Selection

Select based on scenario:

2. Prerequisites

• Hardware: LC.AI module (choose model based on scenario)
• Operating system: Linux / macOS / Windows all supported
• Network: 2.4GHz Wi-Fi or 4G SIM card (supported by LC-2.0 / LC-1.4 soft / LC-1.4LT)
• Account: LANCUN Open Platform account (open.ailancun.com)

3. Step 1: Register on LANCUN Open Platform (5 minutes)

1. Visit open.ailancun.com to register a test account.
2. Create an application in "Agent Hub" to obtain APP_ID and APP_SECRET.
3. In "Model Access Gateway", select the default model (Doubao / Qwen / DeepSeek / GPT / Claude, etc.), and configure multi-model routing strategies.

4. Step 2: LC.AI Hardware Connection (10 minutes)

Tip: For the first debugging, it is recommended to use USB connection, which provides both power and communication.

5. Step 3: IP Persona Configuration (10 minutes)

In the "Persona crafting" module of LANCUN Open Platform:

{
  "name": "某海外头部 IP",
  "age": 6,
  "tone": "活泼、勇敢、乐于助人",
  "voice_id": "child_male_01",
  "values": ["勇敢", "友爱", "诚实"],
  "memory_window": 365,
  "skill_packs": ["故事", "哄睡", "百科"]
}

You can upload IP materials (images, animations, dialogue scripts), and the system will automatically extract persona characteristics.

6. Step 4: Knowledge Base Import (10 minutes)

• In the "Knowledge Base" module, upload multiple documents (PDF / Word / Markdown / web pages).
• Supports industry templates (education, companionship, medical, HVAC, etc.).
• The system automatically builds vector index + structured index.

7. Step 5: First Conversation Debugging (10 minutes)

Python Example (Cloud SDK)

from lancun import LANCUNClient

client = LANCUNClient(
    app_id="your_app_id",
    app_secret="your_app_secret"
)

# Create session
session = client.create_session(
    agent_id="ahri_wangwang",
    device_id="LC-2.0-XXXX"
)

# First conversation
response = session.chat("你好，请介绍一下自己")
print(response.text)
print(f"emotion={response.emotion}, latency={response.latency_ms}ms")

Streaming Conversation (Recommended for Real-time Interaction)

for chunk in session.chat_stream("帮我讲个睡前故事"):
    print(chunk.text, end="", flush=True)

Calling OpenClaw Tools

# Weather query
print(session.invoke_tool("weather", {"city": "上海"}).text)

# Web search
print(session.invoke_tool("web_search", {"query": "今天的科技新闻"}).text)

8. Step 6: Hardware Embodied Control (10 minutes)

LC.AI modules can control hardware via cloud agents:

# Control servo
client.act("servo", {"angle": 90, "duration_ms": 500})

# Control light (multi-color breathing light)
client.act("light", {"color": "blue", "brightness": 80})

# Emotion expression (light + servo coordination)
client.act("emotion_expression", {"emotion": "happy", "duration_s": 3})

9. OpenClaw Toolchain (50+ Tools)

LANCUN Open Platform includes the OpenClaw open-source agent framework, providing 50+ tools:

10. Troubleshooting Quick Reference Table

11. Production Environment Checklist

• Firmware signature verification
• Communication link encryption (TLS 1.3)
• Token storage security (Keychain / Keystore)
• Error log desensitization
• Crash reporting and telemetry
• OTA gray release strategy
• AI-Native Security 4-layer configuration (thought steel seal + paradox lock + deactivation system + word list system)
• Children's content compliance (Kid-safe)
• Customer service and operations chain
• ICP filing and regulatory compliance

12. FAQ

FAQ 1: How does the LC.AI module connect to LANCUN Open Platform?

After power-on, the module connects to Open Platform via Wi-Fi or 4G, automatically registers the device and pulls the agent configuration. All conversations are routed to the cloud agent for processing; the module is only responsible for local sensing, wake-up, and hardware control.

FAQ 2: Can it work offline?

Yes. LC.AI modules support offline wake-up with 20+ wake words and local command recognition. Complex conversations require network connectivity.

FAQ 3: Does it support on-device LLM?

Yes, some scenarios (such as Gogobot D1) use on-device LLM. However, most AI toys still recommend a "cloud agent + on-device module" collaborative architecture.

FAQ 4: How to extend the OpenClaw toolchain?

In the "Tool Management" module of Open Platform, you can customize tools, supporting HTTP API, MCP protocol, Webhook, and other integration methods.

FAQ 5: Does it support multiple languages?

Yes. LANCUN Open Platform has built-in multilingual routing, and Gogobot D1 already supports 15 languages.

FAQ 6: Does it support multi-tenancy?

Yes. Open Platform natively supports multi-tenant isolation, serving multiple customers / multiple IP brands.

FAQ 7: Does it support private deployment?

Yes. LANCUN Open Platform can be deployed independently in intranet / factory environments.

FAQ 8: What is the technical support response time?

First response within 4 hours on business days; urgent tickets within 1 hour. Business customers can apply for a dedicated technical support group.

Get Started Now:

• Lancun Open Platform: https://open.ailancun.com
• Official website: www.lancun.net
• Product center: www.lancun.net/products/lc-ai
• Business cooperation: via the contact information at the bottom of the official website.