How Tencent’s Large Language Models Boost Business with RAG, GraphRAG, and AI Agents

Introduction

In this article we explore Tencent's large language model (LLM) applications across multiple business scenarios, focusing on how cutting‑edge techniques improve model intelligence and user experience. We introduce the model’s broad use cases—content generation, intelligent customer service, and role‑playing—and analyze Retrieval‑Augmented Generation (RAG), GraphRAG, and Agent technologies.

Main Topics

Tencent LLM application scenarios

RAG technology principles and practical use

GraphRAG in role‑playing scenarios

Agent technology principles and applications

Q&A session

Core Application Scenarios

Tencent’s LLMs are applied in many domains such as the WeChat ecosystem, social content, video news, office documents, and games, driving smarter and more efficient services.

Content generation: advertising copy, comment assistance, etc.

Content understanding: text moderation, fraud detection, and similar tasks.

Intelligent customer service: knowledge‑base Q&A, user guidance, etc.

Development Copilot: automated code review, test‑case generation, and related assistance.

Role‑playing: intelligent NPC interactions in game scenarios.

Large Model Application Technologies

Tencent primarily employs three technical approaches for LLM applications:

(1) SFT (Supervised Fine‑Tuning)

Fine‑tunes a base model with domain‑specific data, embedding business knowledge directly into the model for targeted task handling.

(2) RAG (Retrieval‑Augmented Generation)

Combines external knowledge bases and retrieval mechanisms with generation, enhancing explainability and reducing hallucinations; typical use cases include intelligent customer service and document assistants.

(3) Agent (Intelligent Agent)

Leverages external tools to enable the model to perform multi‑step reasoning, planning, and execution, suitable for complex tasks that require sequential decision making.