# Wisdom-Driven Knowledge Augmented Generation at Scale

Chin Keong Lam

AI Engineer World's Fair 2025

[Talk page](https://ai.engineer/talks/9AQOvT8LnMI-wisdom-driven-knowledge-augmented-generation) · [Video](https://www.youtube.com/watch?v=9AQOvT8LnMI)

## Wisdom-Driven Knowledge Augmented Generation at Scale

Chin Keong Lam connects expert decision patterns, a shared knowledge graph and supervised agents to turn retrieved information into situation-aware advice.

Familiarity with retrieval-augmented generation, language-model agents and basic graph concepts will help; no prior n8n or Cypher experience is required.

### When the client needs advice

[0:16](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=16s)

What does an AI system need when a client wants research and advice in a specialist domain, rather than information retrieved from a database? That is the problem behind Chin Keong Lam’s work at Patho.ai. Introducing himself as its founder and CEO, Lam describes starting two years earlier with an invitation through the National Science Foundation’s SBIR program to investigate LLM embeddings. An initial application in drug discovery led to building expert AI systems for corporate clients. Their requests increasingly required the system to perform advisory work.

The starting point is a broader definition of **knowledge**: understanding acquired through experience, education, facts and principles. A knowledge graph connects that material through explicit relationships. In Lam’s framing, the graph should preserve both a domain’s taxonomy and something of its experts’ thought process—the connections they use to reach a decision.

**Knowledge-Augmented Generation (KAG)** is his term for integrating that structured graph with a language model. The intended step beyond basic retrieval is to supply relationships and context that help the model synthesize advice. Lam characterizes this as understanding rather than merely retrieving; the engineering question is how to represent the expert’s reasoning and make it available at the right moment.

![Slide titled “What is KAG?” with a KAG-versus-RAG illustration and a definition below.](https://ai.engineer/api/talk-essay-media/8eb7ea619bf0e1c474be04121ab7b887ada8bb07b07ea2c69c722ce30424ed75)

The introductory KAG slide contrasts knowledge graphs with RAG.

### Represent the decision process, then close the loop

[2:21](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=141s)

Interviews with clients revealed recurring patterns in how experts think and decide. Lam represents those patterns as a graph, or state diagram, with **wisdom** at its center. Wisdom guides the blue decision-making node, which analyzes the green situation node. Decisions therefore depend on the circumstances in which they will be applied. A gold relationship feeds knowledge back into wisdom.

Books, encyclopedias and Wikipedia can supply knowledge, but ingestion alone does not complete this process. The system must synthesize what it has absorbed. Insight supplies a different input: patterns extracted from disorder. For a client tracking product sentiment, scattered social-media posts and X discussions can reveal how customers perceive its products relative to competitors.

The distinction between knowledge, experience and insight is easier to see in Lam’s pizza analogy:

| Input | What it contributes | Pizza example |
|---|---|---|
| Knowledge | What is known | The recipe |
| Experience | What happened before | This oven burns the crust |
| Insight | What to try next | Add honey to encourage caramelization |

These inputs become useful together: the recipe is interpreted in light of a particular oven, and an idea for the next attempt emerges from that context.

**The feedback loop is the essential part of the graph.** A situation informs future wisdom; experience deepens it; insight sharpens it. Lam compares this accumulation to a tree growing stronger roots. In leadership, the loop means using feedback to avoid knee-jerk reactions. In personal growth, it means learning from past mistakes. Wisdom, in this model, is a muscle exercised through repeated use rather than a trophy acquired once.

![Colored nodes joined by labeled arrows under “Storing Wisdom - Knowledge Graph is a perfect fit.” Wisdom sits above the other nodes, with Situation below.](https://ai.engineer/api/talk-essay-media/b9bc3a79241bd00fde94e86f9817a7290ede29cbd1883d7c36fdd599cc5ee9c3)

A graph connects wisdom, knowledge, experience, insight, decision making, and situation.

### Map the graph to a marketing strategist’s work

[6:19](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=379s)

The concrete client application is competitive analysis, work previously handled by a marketing department. The client wants an AI system that can answer a strategic question: how can we win against a competitor in this market? Returning relevant passages is only one ingredient. The answer must connect evidence about the market with the company’s history and present position.

Lam retains the same taxonomy and assigns each part a business responsibility. The wisdom engine becomes an **orchestration agent** that decides what the LLM can see in the current situation and what should happen next.

| Wisdom graph component | Competitive-analysis responsibility |
|---|---|
| Wisdom | Orchestrate context and next actions |
| Decision-making | Generate strategy |
| Knowledge | Supply market data |
| Experience | Supply past marketing campaigns |
| Insight | Supply stored industry insights |
| Situation | Establish sales performance and competitor awareness |

The situation node makes advice conditional on how the client’s products are actually doing. Market information becomes more useful when the model can relate it to that current position.

### Implement the state diagram as supervised agents

[8:30](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=510s)

Lam prototypes the workflow in n8n. It recalls his earlier IoT work with [Node-RED](https://nodered.org/about/), the visual flow tool that originated at IBM and runs on Node.js. The appeal is a visual interface for connecting operations while code performs the work underneath. Community nodes provide building blocks for turning the state diagram into an executable workflow.

The [n8n AI Agent node](https://github.com/n8n-io/n8n-docs/blob/main/docs/integrations/builtin/cluster-nodes/root-nodes/n8n-nodes-langchain.agent/README.md) is central to this prototype. Lam uses n8n for its flexibility; a lighter implementation using something such as LangChain is a possible later choice, not a migration he demonstrates. The linked documentation describes the current Tools Agent behavior, while the recording does not identify its n8n version.

The wisdom engine becomes a supervisory wisdom agent, with options to use models from OpenAI or Anthropic, or an on-premises model. It oversees specialist agents that carry out the responsibilities in the state diagram. Model choice and workflow responsibility are separate: the supervisor’s job remains coordinating the work even when its underlying model changes.

For example, an insight agent visits social media, collects sentiment about the client’s products and writes its findings into the insight portion of a **centralized knowledge graph**. Other agents update their corresponding portions of that same graph. Their outputs accumulate in a unified taxonomy intended to reflect how a marketing strategist organizes evidence.

A human team might keep similar decision material in SharePoint or a collection of folders. Here, explicit relationships connect it for subsequent reasoning. Lam acknowledges that the final answer depends on the model, but places particular weight on the taxonomy and graph structure: the advisor needs the right organization of evidence to make a useful decision. That leads to a closer examination of the competitive-analysis component.

### What a graph contributes to competitive analysis

[11:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=694s)

Why use a knowledge graph, with a database such as Neo4j, for this work? Lam gives five reasons for preferring it to traditional vector RAG in his competitive-analysis application:

- **Relationship context.** Explicit connections between entities can expose gaps in competitor awareness and preserve nuances that matter to a strategy.
- **Accuracy and relevance.** Structured data and semantic relationships can narrow the evidence used in an answer and reduce noise. For a marketing advisor, inaccurate data can lead to bad decisions—and, as Lam emphasizes, cost a contractor the client.
- **Scalability and flexibility.** New sources and relationships can enrich the system over time, provided the taxonomy remains suitable.
- **Rich queries.** Traversing several relationships supports multi-hop questions whose answers require connecting multiple entities.
- **Data integration.** Pictures, graphics and videos can contribute information through LLM and OCR capabilities, with structured and semi-structured results organized in the graph.

These are the architectural benefits Lam seeks. The common dependency is a graph structure that preserves the relationships relevant to the client’s decisions.

### Retrieve the inputs, then calculate

[13:58](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=838s)

The numerical example makes the distinction concrete. Vector retrieval excels at finding semantically similar passages; it does not itself perform the calculation a market analyst needs. Lam begins with a question about Apple’s revenue in 2022. A retrieval system may return a paragraph discussing the revenue, leaving the required numerical work unresolved.

The graph-driven path separates finding evidence from computing an answer:

1. Store the financial data as structured records in an Apple financial knowledge graph.
2. Query the revenue figures for 2021 and 2022.
3. Pass the selected figures to a function that performs the calculation.
4. Return the quantitative result for the marketing decision.

Lam gives 15.23 as the illustrative function output. The slide shows a percentage-change calculation, but the example does not establish the inputs and units behind that number; it should not be read as Apple’s verified fiscal revenue growth. The mechanism is the useful distinction: **structured retrieval supplies the operands, and a function supplies the arithmetic**.

![Slide titled “Knowledge Graph Driven RAG Answer to same question,” with numbered explanation steps, a percentage-change formula, and an example output box.](https://ai.engineer/api/talk-essay-media/40d88c2e5a2bd9beaa973b38a87b43cb8a3f2d6d4f563bcd4cbb751c586f2f71)

The graph-driven answer slide shows a percentage-change calculation and example output.

That separation gives the advisor numerical evidence rather than only a passage that happens to discuss the subject. For decisions based on financial or market quantities, retrieving relevant text and producing the requested calculation are distinct responsibilities.

### Choose the architecture from the question

[15:38](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=938s)

The choice is not an all-or-nothing replacement of RAG. Lam points to LangChain plus Chroma as one way to build retrieval, and notes that it can be combined with a knowledge graph. His architecture slide highlights the wisdom graph he adopts in red. The appropriate combination depends on what the client asks the system to do.

| Client question | Suggested architecture |
|---|---|
| Find product information | Chroma DB with an LLM agent |
| Develop a competitive strategy from current market share | Graph database, Cypher queries and repeated multi-hop retrieval |

The second question requires the system to connect the client’s present position to other evidence before proposing a strategy. Lam describes several loops of querying, rather than one retrieval followed immediately by an answer.

A [Cypher pattern match](https://neo4j.com/docs/cypher-manual/current/clauses/match/) makes the structured-retrieval part explicit. Using an illustrative schema for the Apple example—`Company` nodes connected to `FinancialYear` records—the query can select the two operands without asking the model to recover them from prose:

```cypher
MATCH (company:Company {name: 'Apple'})
      -[:HAS_FINANCIALS]->(financials:FinancialYear)
WHERE financials.year IN [2021, 2022]
RETURN financials.year AS year,
       financials.revenue AS revenue,
       financials.currency AS currency,
       financials.unit AS unit
ORDER BY year;
```

The schema names illustrate the query boundary, not the client’s database schema. Returning currency and unit alongside revenue keeps the quantities interpretable before a calculation consumes them. A strategic advisor needs further queries and synthesis beyond this small retrieval step; Lam expects the repeated graph-query process to improve the resulting advice.

### Extract automatically, refine with the expert

[16:35](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=995s)

Building the graph introduces another choice: construct it manually, use a fully automated LLM graph transformer, or combine the two. Lam recommends the **hybrid approach**:

1. Use an LLM to extract an initial graph.
2. Interview the domain expert to establish the taxonomy the system should use.
3. Prune the graph by removing relationships that do not belong in that taxonomy.

The expert’s contribution is therefore not limited to supplying documents. It also determines which extracted connections should survive into the representation used for advice.

Lam reports 91% accuracy and 85% flexibility in his benchmark. He attributes the accuracy result to extracting structure, then names deterministic reproducibility, traceability and scalability as additional evaluation dimensions. The talk does not define the benchmark task, dataset, baseline, scoring methods or conditions, so the percentages cannot establish an expected improvement for another application.

### The graph is a starting point for continued refinement

[17:41](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1061s)

The ambition is to improve quantitative reasoning and responses to complex questions by giving the system a structured store of expertise. With the right framing, Lam suggests, accumulated knowledge, experience and insight could eventually support capabilities beyond those of the initial expert. He presents that as a possibility, not a demonstrated outcome. A brief “Talk to Jesus” notebook aside follows the conclusion before he turns to the practical starting point.

Lam closes by pointing to a GitHub LLM GraphRAG stack that he describes as sponsored by Neo4j. His proposed handoff is straightforward: start it with Docker, convert text into a graph, then begin pruning. Automatic extraction gets the material into connected form; refining those connections is how the builder shapes it into the domain representation the advisory system needs.

## Resources

- [About Node-RED](https://nodered.org/about/) — The origins and architecture of Node-RED, including its visual flow editor and Node.js runtime.
- [n8n AI Agent Node](https://github.com/n8n-io/n8n-docs/blob/main/docs/integrations/builtin/cluster-nodes/root-nodes/n8n-nodes-langchain.agent/README.md) — Current official instructions for connecting a chat model and tools to n8n's AI Agent node, with version-related behavior notes.
- [Cypher Pattern Matching](https://neo4j.com/docs/cypher-manual/current/clauses/match/) — Examples of querying nodes and relationships using Cypher's MATCH clause.
- [Neo4j Knowledge Graph Builder](https://github.com/neo4j-labs/llm-graph-builder) — Build a Neo4j knowledge graph from unstructured documents using LLM extraction, configurable schemas and Docker Compose deployment.
- [Apple's Fiscal 2022 Results](https://www.apple.com/newsroom/2022/10/apple-reports-fourth-quarter-results/) — Apple's October 2022 announcement reports annual revenue of $394.3 billion and 8% year-over-year growth.

## Slides

### What is knowledge?

Chin Keong Lam opens with a broad definition of knowledge: understanding and awareness gained through experience, education, and comprehension of facts and principles.

[0:32](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=32s) [A definition of knowledge alongside an illustration of a philosopher.](https://assets.ai.engineer/sites/aiecode2025/sha256/5a4d6bf04b120c42f12582987431b4ea695afd52cd3ed8043e998454093df7f2.webp)

What is Knowledge

Knowledge is the understanding and awareness gained through experience, education, and the comprehension of facts and principles.

Patho.Ai at Ai Engineer WorldFair
1

Interface text:
View  Zoom  146%  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document

Speaker notes: Patho.ai started 2 years ago with a invitation from national science foundation for a SBIR grant funding investigating LLM and BERT Driven Drug Discovery Application, since then we also branch out to leverage what we learned building Ai system for Large Corporations. We are currently building Expert Ai system [illegible] clients. Currently, the system we build goes beyond Basic RAG system. Many of our clients are asking for ai system that perform tasks like Research and Advisory roles based on their Area of Interest. Today the talk is about sharing with our fellow Ai engineers what we learned so far building these kind

### What is a knowledge graph?

The speaker presents a knowledge graph as a network of relationships that preserves domain expertise and thought processes, intended to support AI advice rather than database retrieval alone.

[1:20](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=80s) [A definition of a knowledge graph alongside an illustration of a connected brain.](https://assets.ai.engineer/sites/aiecode2025/sha256/7bf332d9e4ef022a73db407dbd199fafd33ee30b0a853de3c2fb7adc6c62a964.webp)

What is Knowledge graph?

A knowledge graph is a systematic method of preserving wisdom by creating a network of interconnected relationships. This graph represents the thought processes and comprehensive taxonomy of a specific domain of expertise.

Patho.Ai at Ai Engineer WorldFair
2

Interface text:
View
146%
Zoom
Add Slide
Play
Table
Chart
Text
Shape
Media
Comment
Share
Format
Animate
Document

### What is KAG?

The slide presents KAG as using structured knowledge graphs to enhance language model responses, contrasting an interconnected graph with RAG’s database retrieval.

[1:53](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=113s) [A slide describes knowledge augment generation and contrasts a knowledge graph with retrieval-augmented generation.](https://assets.ai.engineer/sites/aiecode2025/sha256/ca9b251b0d67f2493eec6f07028e63d4f928f085eed4536c9b9b3264c04efba8.webp)

What is KAG?

Knowledge Augment Generation
Retrieval-Augmented (RAG)
KAG vs RAG
KAG doesn’t just retrieve, it *understands*

KAG (Knowledge Augment Generation) enhances language models by integrating structured knowledge graphs for more accurate and insightful responses, making it a smarter, more structured approach than simple Retrieval-Augmented Generation (RAG) - KAG doesn't just retrieve, it understands.

Patho.Ai at Ai Engineer WorldFair
3

Interface text:
View
146%
Zoom
Add Slide
Play
Table
Chart
Text
Shape
Media
Comment
Share
Format
Animate
Document

### Storing wisdom in a knowledge graph

The graph puts wisdom at the center: knowledge, experience, insight and the situation inform it. Wisdom guides decision-making, which analyzes the situation.

[2:20](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=140s) [A graph connects wisdom with decision making, knowledge, experience, insight, and situation.](https://assets.ai.engineer/sites/aiecode2025/sha256/c1a91865f9a19e3ad1c34e2d9fe7bedaf81924a4d64e843776cdb71478cbdada.webp)

Storing Wisdom - Knowledge Graph is a perfect fit.

Wisdom is like being the smartest player in the game— it doesn’t just react, it finds the winning move!

Wisdom
Decision Making
Knowledge
Experience
Insight
Situation

Relationship labels: guides; synthesizes; informs; reflects on; informs; derives; informs; informs; analyzes

“Knowledge is a GPS showing roads; wisdom is the driver choosing the fastest route based on traffic, weather, and your urgency.”

AIE

Interface text:
View  Zoom  146%  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document

Presenter notes: After interviewing many of my clients who are also expert is certain area of skill, I found that there are common way their thinking, decision making process, the way that make them expert in their area. Knowledge graph seem like a perfect fit to structure to store the wisdom. Here is the Graph or State diagram (if [illegible]
[illegible] as you can see
[illegible] isn’t just knowledge—it’s a dynamic system where experience, insight, and action feed each other. Let me show you how this works.
[illegible] each node and relationship (use the graph as visual aid):
[illegible] (Purple Circle))
[illegible] core—Wisdom isn’t static. It actively guides decisions and is fueled by other elements.
[illegible] Output: Decision Making ((Blue Circle))
[illegible] isn’t passive. It guides decisions, helping us choose wisely.

### An advisor for competitive analysis

The proposed wisdom engine guides a strategy generator and connects market data, past campaigns, industry insights, and competitor weaknesses. The speaker describes it as an orchestration agent that advises what the LLM sees.

[6:47](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=407s) [A wisdom engine graph connects a strategy generator, market data, past campaigns, industry insights, and competitor weaknesses.](https://assets.ai.engineer/sites/aiecode2025/sha256/6f11f0b0ba60ca46766423d55956bb9b7dcb4626fd4cbf82cc03f837e967d637.webp)

Not Just a Chat Bot - it is a Advisor? For Competitive Analysis Application.

Wisdom Engine
Strategy Generator
Market Data
HP’s Past Campaigns
Industry Insights
Competitor Weaknesses

Relationship labels: guides; synthesizes; informs; reflects on; informs; derives; informs; informs; analyzes

“What if your chatbot didn’t just analyze competitors—but outthought them? Meet our Wisdom Graph-powered AI, designed to turn data into strategic dominance.”

AIE

Interface text:
View  Zoom  146%  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document

Speaker notes:
This is one of chatbot that behave like a Advisor.This chatbot didn’t just analyze competitors—but outthought them? Meet our Wisdom Graph-powered AI, designed to turn data into strategic dominance.”
[illegible] the bot as a “strategic advisor” (not just a data tool).
[illegible] its feedback loops (learns from every analysis to get smarter).
[illegible] “Synthesizes”
[illegible] means:

### Why use a knowledge graph for competitive analysis?

The speaker argues that knowledge graphs can improve competitive analysis by connecting entities, supporting multi-hop queries, and integrating diverse data sources. He highlights accuracy for marketing decisions.

[11:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=694s) [Five proposed benefits include contextual understanding, accuracy, flexibility, rich queries, and data integration.](https://assets.ai.engineer/sites/aiecode2025/sha256/da95ddbc31d8873bb2908765846788ea5cd2333d39b08af7338df1aea20d6166.webp)

Competitive Analysis , why use Knowledge Graph?

1.Enhanced Contextual Understanding: Knowledge Graph-based RAG (Retrieval-Augmented Generation) systems excel at capturing and representing complex relationships between entities. This leads to a deeper contextual understanding, which is crucial for competitive analysis where nuanced insights can make a significant difference.
2.Improved Accuracy and Relevance: By leveraging structured data and semantic relationships, Knowledge Graphs can provide more accurate and relevant information retrieval compared to traditional vector stores. This ensures that the generated content is not only relevant but also precise, reducing the noise and improving decision-making.
3.Scalability and Flexibility: Knowledge Graphs are inherently scalable and can easily integrate new data sources and relationships. This flexibility allows for continuous improvement and adaptation to changing market conditions, making them a robust choice for long-term competitive analysis.
4.Rich Query Capabilities: Knowledge Graphs support complex queries that can traverse multiple relationships and entities, providing richer and more detailed insights. This is particularly advantageous for competitive analysis, where multi-faceted queries can uncover hidden patterns and trends.
5.Enhanced Data Integration: Knowledge Graphs can seamlessly integrate diverse data sources, including structured, semi-structured, and unstructured data. This holistic approach ensures a comprehensive view of the competitive landscape, enabling more informed strategic decisions.

AIE
7

Interface text:
View
146%
Zoom
Add Slide
Play
Table
Chart
Text
Shape
Media
Comment
Share
Format
Animate
Document
What is Knowledge graph?
What is KAG?

### Knowledge graphs versus vector stores

The speaker argues that vector RAG can return similar passages when market analysis needs numbers. Knowledge graphs organize data and relationships to support calculations and aggregation.

[13:58](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=838s) [A slide contrasts numerical reasoning and aggregation limitations with structured data and complex queries.](https://assets.ai.engineer/sites/aiecode2025/sha256/e52c0b91b5acde06690cea8568fc8c1f591fcf4bdeccda4eea2128e02a19ac91.webp)

Knowledge Graphs vs. Vector Stores: A Quantitative Analysis Example

Problem: Limitations of Vector Store-Driven RAG in Quantitative Analysis

• Limited Numerical Reasoning: Vector stores excel at semantic similarity, but struggle with complex numerical calculations and logical reasoning.
• Inaccurate Data Aggregation: Without a structured understanding of data relationships, vector stores can misinterpret queries, leading to incorrect quantitative results.

Solution: Knowledge Graphs for Enhanced Quantitative Analysis

• Structured Data Representation: Knowledge graphs store data in a structured format, enabling precise queries and calculations.
• Complex Query Capabilities: Knowledge graphs can handle complex queries involving multiple entities, attributes, and numerical operations.
• Accurate Data Aggregation: By understanding the relationships between data points, knowledge graphs can accurately aggregate and analyze data.

AIE
8

Interface text:
View  Zoom  146%  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document
Here is a slide showing the limitation of vector store Driven RAG especially in Quantitative analysis. Firstly, traditional RAG have ....

### Vector-search RAG answer example

The example illustrates how a question about Apple’s 2022 revenue can return qualitative passages about products and performance without a numerical answer.

[14:40](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=880s) [Five example passages discuss Apple's performance instead of providing a numerical revenue answer.](https://assets.ai.engineer/sites/aiecode2025/sha256/6c2c5c38e28e57055f8ccbfd3b864b5bb299cb347e8949babffeb3678490b08c.webp)

Normal RAG (Based on Vector Search Alone Rag Answers)

• Passage 1: “Apple reported another record-breaking year in 2022, with strong performance across all product categories. The iPhone 14 series was particularly popular, driving significant revenue growth.”
• Passage 2: “Despite supply chain challenges and economic uncertainty, Apple managed to maintain its position as the world's most valuable company. The strong demand for its products, especially the iPhone and Mac, contributed to its impressive financial performance.”
• Passage 3: “Apple's services business continued to grow rapidly in 2022, driven by the success of the App Store, Apple Music, and Apple TV+. This segment is becoming an increasingly important revenue driver for the company.”
• Passage 4: “The global chip shortage impacted the production of certain Apple products, but the company was able to mitigate the impact through strategic sourcing and inventory management. However, it did affect overall revenue to some extent.”
• Passage 5: “Analysts are optimistic about Apple's future prospects, citing the growing popularity of its products, especially in emerging markets. The company is also investing heavily in new technologies such as augmented reality and autonomous vehicles, which could drive significant growth in the coming years.”

AIE
10

Interface text:
View · 146% · Zoom · Add Slide · Play · Table · Chart · Text · Shape · Media · Comment · Share · Format · Animate · Document
What is Knowledge graph?
What is KAG?
Here is a slide elaborating the output from Traditional RAG answer, long passage without any Business Driven Quantitive Answer. All because of the nature of the process where traditional RAG chunk up the initial data and stored them without any indexing or relationship.

### Knowledge-graph RAG answer example

The example queries structured revenue figures for 2021 and 2022 and calculates a 15.23% change in Apple’s revenue.

[14:43](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=883s) [An example calculates a revenue percentage change and displays an illustrative result of 15.23%.](https://assets.ai.engineer/sites/aiecode2025/sha256/d7e02e948fab334284df3700a39c0bd5fd1314c2dd157a0ea457fea13a25e797.webp)

Knowledge Graph Driven RAG Answer to same question

Explanation:
1. Data Source: Assumes a knowledge graph named Apple_Financial_Data.
2. Query: Selects the revenue figures for 2021 and 2022.
3. Calculation: Calculates the percentage change using the formula:
   Percentage Change = ((New Value - Old Value) / Old Value) * 100
4. Result: Returns a single value representing the percentage change in Apple's revenue from 2021 to 2022.

Example Output:
Percentage_Change
15.23%
Export to Sheets

This query directly accesses the stored numerical data in the knowledge graph, providing a precise and accurate answer to the query “What was the percentage change in Apple's revenue from 2021 to 2022?”

AIE
11

Interface text:
View
146%
Zoom
Add Slide
Play
Table
Chart
Text
Shape
Media
Comment
Share
Format
Animate
Document

### RAG and knowledge graph solutions

The speaker suggests Chroma and an LLM agent for simple product queries, and a knowledge graph with GraphDB and Cypher for harder questions, using multi-hop queries with Q&A and RAG.

[15:30](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=930s) [Diagram connecting data sources, language models, databases and knowledge graphs to Q&A and RAG.](https://assets.ai.engineer/sites/aiecode2025/sha256/586ba91153a5b9b09086cb4b49b9f2495f902b8463370ff33838839e41a9d256.webp)

So Many Solutions, FOCUS On Red Solution

SQL
Excel
CSV
PDF
W
txt

SQL LLM Agents
+ SQL DB + SQL query

SQL LLM Agents +
+ SQL DB + SQL query

LLM + VectorDB +
Embedding + Vector search

LLM Agents + Knowledge Graph
+ GraphDB + Cypher query

LLM + Knowledge Graph
+ GraphDB + Embedding + Vector search + Cypher query

RAG-GPT
LLM + Chunking + VectorDB
+ Embedding + Vector search

Knowledge Graph + LLM +
Chunking + embedding +
Vector search + GraphDB

Microsoft project
LLM Agents
+ Chunking + Knowledge Graph + GraphDB + Cypher

Q&A
RAG

Microsoft Azure
LangChain
neo4j
Chroma
SQLite

AIE
Source:
12

Interface text:
146%
View  Zoom  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document

Presenter notes: This slide show RAG and KAG can be build using many different Vendor. The one I adopted for Wisdom Graph is in RED. Normally you can see if client just asking for simple rag that can perform product information query, you can just use simple chroma DB with LLM Agents will cover it. If they start for ask [illegible] question involving like “How can I beat my competition based on my current market share?” Than Notice the Texts highlighted in RED, I would recommend you build start interviewing their in house Marketing department and collect information based on the taxonomy of Wisdom Graph I shown [illegible].

### Constructing a knowledge graph

The speaker recommends a hybrid: use an LLM to extract a graph, then interview a domain expert to build a taxonomy and prune relationships. The slide contrasts manual consistency with the ease and inconsistency of LLM extraction.

[16:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=994s) [Comparison of domain expert, hybrid and LLM approaches, with advantages and disadvantages.](https://assets.ai.engineer/sites/aiecode2025/sha256/2a33eeec017370ac1fe40c7232dd11b326283f862f397cec1b77ff1cf056e201.webp)

How to construct the knowledge graph?

Domain expert
Pros: Consistent graph structure
Cons: Not easy to implement

Hybrid: Use LLM’s knowledge (e.g: GPT 4) to guide you in constructing the Graph Knowledge)
Pros:
- Consistent graph structure.
- Less expertise are needed in collaboration with LLM.
Cons: Still requires some basic knowledge of coding and graph knowledge implementation

LLM (e.g: LLMGraphTransformer)
Pros: Easy to use
Cons: Inconsistent graph structure (due to nondeterministic behavior of the LLMs)

No matter which approach we choose, after constructing the knowledge graph, we need to populate the GraphDB by defining the desired nodes and relationships ourselves. Therefore, this inconsistency will mostly affect the generated information rather than the structure of our GraphDB.

Interface text:
View  Zoom  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document
146%

Speaker notes: This slide show the Different method of grounding your Truth with Knowledge graph, on the left it is graph manually built by interviewing the expert and probing his thought process just like the wisdom Graph I have shown. This is timing consuming but the graph structure after you built is very consistent and on the [illegible] this is using LLM Graph Transformer available in python library, this is extremely easy, you just ingest your data and pray that the LLM you paid can understand and build the graph correctly for you which normally will not be the case. Inconsistency graph will result in non deterministic behavior. What I [illegible] the method in the middle based in hybrid method, where you used LLM to build you initial graph and interview the expert in that space by pruning the graph. This have pro of reducing your time consuming graph building and can require Less Expertise, all you need to is to interview the expert and [illegible] graph as he give you feedback.

### Benchmark results

In a benchmark described as using more than 1,000 questions, GraphRAG scores 91% versus 58% on QA accuracy but takes 165 ms versus 105 ms to respond.

[17:13](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1033s) [Table comparing reported RAG and GraphRAG results across accuracy, flexibility, reproducibility, traceability, response time, statistical queries and scalability.](https://assets.ai.engineer/sites/aiecode2025/sha256/30f9cf6f44627f2a039b006bbd141f715ad42cc12c24b1719182e4808a8d204c.webp)

Benchmark Result

Key Benchmark Metrics Table

Metric | Normal RAG | GraphRAG (Hewy) | Superior System | Reason for Superiority
Accuracy (QA Score) | 58% | 91% | GraphRAG | Ontology extracts structured specs (e.g., blade counts from PDF tables).
Flexibility | 40% | 85% | GraphRAG | Handles temporal and multi-hop queries across 2019–2025 reports.
Reproducibility | 65% | 97% | GraphRAG | Deterministic graph queries ensure consistency across runs.
Traceability | 25% | 94% | GraphRAG | Answers link to PDF sections and component nodes.
Response Time | 105 ms | 165 ms | Vector RAG | Faster for keyword searches in large text corpora.
Statistical Queries | 15% Success | 93% Success | GraphRAG | Structured properties enable exact numerical retrieval.
Scalability | Moderate | High | GraphRAG | Efficiently maps 10M+ component relationships.

Interface text:
View | Zoom | 146% | Add Slide | Play | Table | Chart | Text | Shape | Media | Comment | Share | Format | Animate | Document

Speaker notes: Here this is the standard way my team used to Measure our AI system performance, if you client ever asked you to show why KAG or GraphRag is better than normal rag, give them this table quantified by the metric as shown. This the result of giving two system more than 1000 questions constituting of Simple [illegible], Temporal Question and Multi Hop Questions.

### Conclusion

The speaker proposes that structured wisdom knowledge graphs improve KAG’s quantitative analysis and complex-query responses. With the right framing, systems could scale, store wisdom, and surpass the initial expert’s intelligence.

[17:40](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1060s) [Conclusion describing the proposed role of wisdom knowledge graphs in quantitative analysis and complex queries.](https://assets.ai.engineer/sites/aiecode2025/sha256/3838a29d017263fbf62fcffdaf8d9e8e293187675346d367c012c05c0346e4f5.webp)

Conclusion

By leveraging the structured nature of wisdom knowledge graphs, we can significantly enhance the quantitative analysis capabilities of KAG systems, enabling more accurate and insightful responses to complex queries. By using the wisdom Driven system as highlighted, together we can build smarter ai system that scale and store wisdom with the right framing potentially surpassing the intelligence of the initial Expert it meant to serve.

https://www.linkedin.com/pulse/just-aware-you-talked-oncologist-chin-keong-lam-dzore/
AIE
19

Interface text:
View
146%
Zoom
Add Slide
Play
Table
Chart
Text
Shape
Media
Comment
Share
Format
Animate
Document

### Neo4j booth

Alongside a photo from the Neo4j booth, the speaker recommends a Neo4j-sponsored LLM GraphRAG stack for converting text into a graph and pruning it.

[18:06](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1086s) [Closing slide showing two people at a conference booth.](https://assets.ai.engineer/sites/aiecode2025/sha256/c7c68692815edb9b84108ea9ee3d820c1aed55d3b3ff2fb8c97840a85f6d2cde.webp)

What will Jesus Do ? - Neo4j Booth AIE

Photo text:
Build GenAI [illegible]
With Graph[illegible]

https://www.linkedin.com/pulse/just-aware-you-talked-oncologist-chin-keong-lam-dzore/

AIE
20

Interface text:
View  Zoom  146%  Add Slide  Play  Table  Chart  Text  Shape  Media  Comment  Share  Format  Animate  Document

## Full transcript

[0:00](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=0s) [upbeat music] So hi.

[0:16](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=16s) Hi, everybody. Uh, my name is Chin Keong Lam. Um, I'm a founder and CEO of Patho.ai. Uh, bit of background about my company. Uh, Patho.ai started two years ago with, uh, invitation from National Science Foundation, from the SBIR grant funding investigating LLM embed.

[0:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=34s) We did, uh, LLM embed-driven, uh, drug discovery application. Uh, since then, we branch out to leverage what we learned about building AI system for large corporation. We are currently building expert AI system for several clients.

[0:49](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=49s) Currently, the system we build goes beyond RAG system. Um, many of our client is asking for AI system that perform tasks like, uh, research and advisory role based on their area of interest.

[1:01](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=61s) Uh, today, the talk is about sharing with our fellow AI engineer what we learned so far building this kind of system. Okay. Uh, what is knowledge? Okay, generally, philosophically, I say, uh, knowledge is the understanding and awareness gained through experience, education, and a comprehension of facts and principle.

[1:20](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=80s) And that lead to the next question, is what is knowledge graph, right? So knowledge graph is a systematic method of preserving wisdom by connecting them and creating a network or interconnect relationship.

[1:32](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=92s) That's important. The graph represent a thought process

[1:36](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=96s) and comprehensive taxonomy of a specific domain of expertise. That's why this is, is very important for people moving forward. It's about AI system that think a lot and return, uh, advice instead of just retrieve, you know, data from your database, right?

[1:53](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=113s) So that comes to the development of this, uh, KAG. Okay, what is KAG? KAG stands for Knowledge-Augmented Generations, and it's different from RAG, okay? It is enhanced language model by integrating structured knowledge graph for more accurate and insightful response, making it smarter, more structured approach than a simple RAG.

[2:14](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=134s) KAG doesn't just retrieve. Remember, it understand. This is different.

[2:21](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=141s) Okay. After inve- interviewing a lot of my client, okay, so, uh, who also are expert in a certain area of skill. I found that there are common ways of their thinking, decision-making process.

[2:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=154s) The way that make them expert in their area, knowledge graph seems to be a perfect fit. So here is a graph or state diagram if you are computer engineering grad like me.

[2:44](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=164s) So, um, it shows wisdom, the f- the wisdom node, as you can see, is the, is the core, right? It's wisdom, it just isn't static. It actively guide decision and fueled by other element.

[3:02](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=182s) The output from the wisdom actually goes to decision-making in the blue, right? Wisdom isn't passive. It guide decision, helping us choose wisely, okay? And then the decision-making analyze the situation given in the circle in the, uh, green.

[3:20](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=200s) And decision aren't make, you know, in a vacuum, okay? They analyze real-world situation. That's the difference, okay? So look at the wisdom input, okay? Look at the relationship feedback from the knowledge to wisdom in gold color.

[3:36](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=216s) Example of that is knowledge to wisdom, like all your book smart and encyclopedia, Wikipedia, whatever you store. Plus, once that data get absorbed by LLM, whatever model you use out there, it need to regurgitate that and understand.

[3:55](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=235s) That's why it's very important that wisdom is able to synthesize the data after you ingested knowledge. You know, that's a kind of abstract, but I'll, I'll come, come to that later, what I'm talking about, okay?

[4:05](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=245s) Okay, from insight. Example of that is wisdom derive pattern from chaos. Like, some of my client has a lot of social media. They, their product, how do they, you know, track their product sentiment from, from social media, right?

[4:19](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=259s) Sometimes it's very chaotic. And from X tweet, right? So, so from that, you can see some pattern of their competitor versus, uh, current what my product is. That, that's like example of that, and I will go to that later.

[4:31](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=271s) Okay. When all these connected nodes matter together, why do they matter? All the nodes relate to one another to ever increase, um, enriching our wisdom storing system. Okay, this talk is about storing wisdom, right?

[4:45](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=285s) So knowledge tells you what it is, right? And experience tell you what worked before. Insight invent what to try next. Right? Like a pizza, knowledge is recipe.

[5:01](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=301s) Experience is knowing your oven burn crust. Insight is like, hey, it is at adding, you know, honey to the crust, it will caramelize perfectly, right? So the most important part of the knowledge graph is feedback loop, okay?

[5:16](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=316s) Feedback isn't one-way street. It learn from itself. Look at the feedback from the, uh, going back to all the node from insight to wisdom, okay? Um, situation inform future wisdom.

[5:29](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=329s) Experience deepen it. Insight sharpen it. Like a tree growing roots. The more it fed, the stronger it get. Now, I want to ask you a question, in general. Where do you see this circle in your life?

[5:44](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=344s) Maybe a tough decision that, you know, taught you something.

[5:50](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=350s) So one practical application for leadership is wisdom. Avoid knee-jerk reaction by learning from feedback. As for personal growth, ever notice how past mistake make you wiser? That's the loop into action.

[6:05](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=365s) All this. So the takeaway from the slide in this is wisdom isn't a trophy you earn. It is a muscle you exercise. The more you feed knowledge, experience, insight, the more that guides you.

[6:19](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=379s) Now, I will show you how it being mapped to my current client. You know, all this is like very abstract, right? So how I-- one of my clients actually doing a competitive analysis, uh, they used to have a marketing department doing that, but they want AI to do that, right?

[6:36](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=396s) They, they asked me to build a system. This exactly what I did with the same taxonomy of storing all this. So this tax-taxonomy will be... Later on, I talk about how multi-agent is going to handle all that.

[6:49](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=409s) Here is one of the chatbot that I built for my client to do, you know, not just some, uh, we-- not just some chatbot, okay? It's our Wisdom Graph-powered AI designed to turn data into strategy, right?

[7:01](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=421s) Dominant. So what kind of question I talk about? Talk about how do I win my competitor in this market space? That's kind of very sophisticated question, right? So without, uh, if, if you do simply just RAG by...

[7:11](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=431s) First speaker talk about RAG, right? So it's not going to cut it. They're not going to able to answer that kind of question, okay? What I did is this.

[7:18](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=438s) Uh, we retain the same taxonomy and, uh, the wisdom is then mapped. The same engine there, the wisdom engine. Wisdom engine is like a orchestration agent that does a lot of decision-making, including advising what the, the LLM is able to see based on the current situation, what to do next, right?

[7:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=454s) So, um, what I did is, uh, for the, uh, decision-making, I map it to a strategy generator. So these customers talk- are talking about a competitive analysis, right? So, um, I map the knowledge.

[7:46](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=466s) In term of knowledge, what do they have? They have market data, right? So I map this experience to HP. It's one of a kind, okay? Past campaign. So they have a lot of campaign, doing a lot of, uh, marketing.

[7:59](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=479s) And then, um, the insight is actually mapped to, uh, in-industrial insight. They have a database doing, storing that. And then, of course, the most important is this, the s-situation.

[8:10](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=490s) The situation is how, how am I doing? How my product selling, right? So, so that, that is like a situation, and then I map that to a competitor awareness.

[8:19](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=499s) That means to say, if you make sh- the LLM aware of that, you probably get a very good answer, and then, you know, the chatbot will probably be doing the right thing, advising.

[8:30](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=510s) So from here, very high level, you know, state diagram, all that. How do I map it to a system that drive? Well, here comes the trick. So anybody here heard of n8n?

[8:42](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=522s) All right. All right. It's all good. So, so I, I first encounter similar situation when my, my past IoT project, which is Node-RED developed by, uh, uh, IBM, right?

[8:53](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=533s) So it's the same kind of thing. It's like no code. But, but underneath the hood, there's a bunch of code, okay? It's all Node.js code, okay? So, uh, but, but for the, for, for proving your concept and all that, it's very, very, very flexible.

[9:05](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=545s) And I, I re-highly recommend that. And, and, and here, here you can take a look at the, the workflow, the workflow. I enable the implementation of this complicated state diagram with, um, uh, what I say is there is a different community node.

[9:18](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=558s) One of the very powerful node is the AI Agent node. Well, previously, n8n is just a workflow automation tool. I'm not selling for n8n here. I'm just telling you I'm using it, uh, for pro-prototyping.

[9:28](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=568s) Uh, further down the road, maybe the client say, "Oh, it's too... Like I, I really need to, you know, go lightweight." Maybe we will switch over to some other LangChain or whatever.

[9:37](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=577s) But, uh, we actually use this. I, I mapped the previous, uh, state diagram from the wisdom engine. I actually mapped that to our, our wisdom agent. Okay, wisdom agent is now have the option to drive a different model, like OpenAI model, Anthropic model, and even on-prem model.

[9:54](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=594s) And then that, the key in making the state, uh, mach- the state machine work is that my wisdom agent is now overseeing, like a supervisory agent, uh, all these other agent that do, uh, whatever I say on the state diagram logic.

[10:07](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=607s) Um, for example, the, uh, state of, uh, going into a node of insight. Insight a-agent will has to do, go to the social media, look for the sentiment of all your product, and then collect that, and then pump that.

[10:22](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=622s) The... You can see that at the dot at bottom that, that we are connected to a, a, a, a, a centralized, uh, graph. The centralized graph will be able to get updated by different agent.

[10:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=634s) Uh, insight agent will update the, their, their perspective, like part of that graph for the, uh, as I say, for this particular, uh, uh, insight node. So, so all the unified knowledge graph will contain the taxonomy that eventually just think like the marketing strategists.

[10:54](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=654s) The way that they, they was probably, if you are doing manually, they probably would think in your, in your SharePoint or all this, you know, folder will store the same kind of, uh, you know, wisdom I call it, to make decision based on that.

[11:07](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=667s) So the, the final decision is LLM also depend on the model that you use. Uh, but I, I, I, I pretty much think that not really the way that I think the final decision come when you make a right decision from the advisor output is basically depend on all the taxonomy, the graph structure.

[11:24](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=684s) That's very important. So come to that, I, I want to go deep down how I implement one of the node, uh, just to go a bit technical on this competitive node, how do I implement that.

[11:34](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=694s) Okay, before I do that, okay, competitive analysis, right? Why, why, why you can actually just use RAG? Why do you want to use a knowledge graph like now Neo4j?

[11:43](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=703s) Well, if you ever been asked that question, tell them these five, uh, five reason. Okay, first reason is knowledge graph, you know, uh, system excel at capturing and representing complex relationship with the entities.

[11:56](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=716s) That is covered by the first speaker, but I'll just reiterate that. This lead to a deeper contextual understanding, which is crucial for competitive analysis, where this, in this case, the nuance insight can be- Signific-- make a significant difference, okay?

[12:09](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=729s) You want to find the gap in your competitor awareness. Now, this is very important. The second is improve accuracy. By leveraging structured data and semantics relationship, knowledge graph can provide more accurate and relevant information compared to traditional vector RAGs.

[12:24](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=744s) Um, this ensure generated content is not only relevant, but also precise and reduce the noise and improve decision-making, making. This, in this case, the bot is supposed to help the guy that is marketing department make decisions.

[12:37](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=757s) So, so you better make this work. Improve accuracy. Any inaccurate data, you will be out of the contract, out of the door, right? So the, the very important, okay, you are talking a contract work like me, I have to make the RAG as accurate as possible.

[12:49](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=769s) So the third is scalability and flexibility. The graphic, you know, knowledge graph are inherently scalable and can integrate to new data source and relationship. The flexibility allow the continuous improvement.

[13:00](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=780s) As I say, if your taxonomy is correct, you will continue to improve and, and enrich, right? So, so that is important. And also, rich query capability. Knowledge graph support complex query, traverse to multiple relationship entity, provide richer and more detailed insight.

[13:16](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=796s) This is particularly advantage for a competitive analysis where multifaceted query like, like what the first speaker said, it is super authoriously good in answering things that normal, normal, uh, normal RAG will fail.

[13:28](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=808s) It's like multi-hop question, okay? This is very important. And then the final one is the enhanced data integration. Uh, knowledge graph can seamlessly integrate diverse data source, pictures, graphics, videos.

[13:42](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=822s) Uh, however it is, now that LLM is so powerful, we have OCR capability, it can do that. As long as you have a right structure of the graph, semi-structure and structure, the holistic approach ensure comprehensive view of the competitive landscape enable more informed strategic decision-making.

[13:58](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=838s) Okay. So one of the-- This is, uh, I'm gonna just very briefly go through this. It's just a, a example of the some of the thing. Like, um, problem of a vector's RAG, you know.

[14:08](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=848s) Vector RAG is really, really bad in answering limited numerical res-reasoning. Vector store excel, you know, at semantic sim-similarity, but struggle with complex numerical calculation. This is why, uh, for, uh, market an-analysis, uh, that I'm building the chatbot for, uh, they actually rely on number instead of just, you know, returning example like this, like kind of...

[14:30](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=870s) If you ask like, uh, what is the Apple, uh, revenue, uh, between, uh, two, uh, you know, what's the revenue in two thousand twenty-two? They, they probably will give you a bunch of these kind of a passage, right?

[14:42](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=882s) Retrieve a paragraph. Instead of, uh, this kind of a very, very precise thing like, uh, the, the answer is, uh, you know, uh, knowl-knowledge graph is able... Because the, uh, the, the, the data is already there in structured form.

[14:54](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=894s) The data source assume a knowledge graph name, this particular, uh, in, in this particular case, Apple financial data. The query will be able-- The query engine will be able to select the, the revenue figure from twenty twenty-one to twenty twenty-two, and, and then do a function call.

[15:09](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=909s) The function call will eventually give, come out with fifteen point two three, which is exactly what the marketing guy was looking for. A very quantitative stuff that most of the decision were based on that because you have the evidence, not just some passage that you retrieve from the data.

[15:23](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=923s) It, it's basically evidence-based decision-making. It's very important for this kind of, uh, complicated RAG system that, you know, uh. So, um, there's a jungle out there right now. You can use different kind of, uh, uh, uh, uh, uh, of thing to build yours, uh, you know.

[15:38](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=938s) Uh, this is just a snapshot of that, you know. You can actually use, uh, LangChain plus Chroma to, to build your own RAG, and then you also can combine that with your knowledge graph.

[15:48](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=948s) Depend on, on, on your use case, okay? If, if, uh, the, this slide show that the RAG and the KAG can be built with many, okay. I adopt that wisdom graph in red color.

[15:59](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=959s) Normally, you will see if client is just asking for a simple RAG that perform product information query, you can just use a simple Chroma DB with LLM agent. And if you start to ask so complicated questions like, "How can I beat my competition based on my current market share?"

[16:15](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=975s) Well, this will be able, the, the, uh, the, the, the thing that I will probably be adopting is knowledge graph here with, uh, graph DB plus Cypher query and it will create Q&A, and also train my RAG to perform several loop of, uh, we call multi-hop query.

[16:32](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=992s) And this probably will give a very good answer.

[16:35](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=995s) So, uh, and then it come to the another question. When I was trying to extract my, uh... Oh, I think my time is, uh, is almost up. Okay. So anyway, this is like to say, uh, the first speaker talk about the extraction, right?

[16:49](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1009s) There's a very simple way to extract. On the right side is, like automated, totally automated LLM graph transformer. On the left is like manual. I would probably rec-recommend the center hybrid model, which is like after you use the LLM to extract your graph, you ask to interview the, the expert that you're gonna build, uh, to, to build

[17:06](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1026s) a taxonomy, right? To prune the graph. We call it pruning a graph. Remove a lot of relationship there. Then that, that will be okay. And, um, I will try to just highlight this.

[17:15](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1035s) This is the result of benchmark that we did. Okay. Anybody ask you, you know, why you want to use graph, right, or KAG? Okay. First is accuracy. I had achieved ninety-one percent because it's really good in extract structure.

[17:26](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1046s) Second is flexibility, eighty-five percent. Third is rebo-ducibility, reproducibility, deterministic. And then the fourth one, traceability. And finally, uh, most important is scalability. So in conclusion,

[17:41](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1061s) by leveraging structured nature of wisdom knowledge graph, we can significantly enhance the quantitative capability of KAG system and enable more accurate and insightful response to com-complex query. By using wisdom-driven system as highlighted, together we can build smarter AI system that can scale and store wisdom with the right framing, potentially surpass the, uh, intelligence of the initial expert

[18:04](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1084s) that we meant to serve. [audience applauding] So we, uh, talk to Jesus, you know. What do Jesus do? Talk to Jesus. He's in, uh, in, in a notebook. This is my good friend.

[18:15](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1095s) And, uh, anybody that want to build graph, we have a good, uh, so-called LLM GraphRAG, a stack on GitHub that, um, is sponsored by Neo4j. And out of the box, just spin up your Docker.

[18:27](https://www.youtube.com/watch?v=9AQOvT8LnMI&t=1107s) The next thing you know, your text is going to be converted to your graph, and you can start happy pruning your graph. Thank you. Thank you so much. [audience applauding] [outro music]
