1 of 23

Build, Launch & Contribute to AI Agents using Gaia

2024

2 of 23

Build, Launch & Contribute to AI Agents using Gaia

👋

Developer Advocate @ Gaia 🌱��Organized over 200 hackathons across the globe for AngelHack��Also occasional hackathon winner 🏆

Harish Kotra

@HarishKotra

3 of 23

How many of you used ChatGPT or Claude?

4 of 23

How many of you know what AI Agents are?

5 of 23

How many of you have used an open-source LLM?

6 of 23

How many of you have tried to create an AI agent?

7 of 23

7

Introduction to Gaia

How to launch an AI Agent

Demo of Gaia

Contributing to Open Source & Gaia

Table of Contents

8 of 23

8

Introduction to Gaia

Gaia turns decentralized AI into a network of living knowledge, evolving in real-time to meet the ever-changing needs of humanity.

Our mission is to unlock living knowledge systems, ensuring these AI services reflect the unique styles, values, knowledge, and expertise of their creators.

9 of 23

9

Getting Started with Gaia

Intro to Gaia

GaiaNet Node Operator: This is a person or organization that runs a Gaia node. They bring their own knowledge and specially-trained AI models to the network.
The node consists of several parts:

An API Server that handles various AI tasks like using tools, calling functions, running AI models (LLM), and retrieving relevant information (RAG).
A system for managing different AI instructions (prompts).
A database for storing information in a way that AI can easily understand and use.
Models for understanding different types of data (text, images, etc.).

At the bottom, there's something called WasmEdge Runtime. This is like a universal translator that allows Gaia to run on almost any type of computer hardware.
Each node gets a unique ID, which connects it to the larger GaiaNet Domains. This is how individual nodes become part of the bigger network.
On the right, there's a Marketplace. This is where people can buy and sell different AI services and components, like custom AI workflows, specialized knowledge, or fine-tuned AI models.

The big idea here is that Gaia allows many different people to contribute their AI expertise and resources to a decentralized network. This network can then offer a wide range of AI services to users, potentially rivaling big, centralized AI companies but in a more open and diverse way.

10 of 23

10

Resources

What can you build

with Gaia?

dApps

Decentralized AI applications

OpenAI Compatible APIs

Drop-in OpenAI replacement

Personalised AI Agents

Custom AI assistants

Digital Twins

API-powered virtual replicas

Plugins & Integrations

Extend web3 + AI capabilities

Decentralised AI Agent Marketplaces

Distributed AI service network

Drop-in OpenAI replacement

Seamlessly integrate with existing OpenAI-based projects
Familiar API structure for quick adoption

Decentralized AI applications

Build AI-powered dApps with enhanced privacy and control
Leverage blockchain technology for transparent and secure AI services

Custom AI assistants

Create tailored AI agents with specialized knowledge and capabilities
Fine-tune models to match specific use cases and industries

API-powered virtual replicas

Develop digital twins with versatile API integration
Enable real-time data synchronization and simulation

Distributed AI service network

Participate in a decentralized marketplace for AI services
Monetize your AI models and knowledge bases

Extend AI capabilities

Enhance AI functionality with custom plugins and integrations
Connect to external tools and data sources for comprehensive solutions

11 of 23

11

Introduction to Gaia

How to launch an AI Agent

Demo of Gaia

Contributing to Open Source & Gaia

Table of Contents

12 of 23

12

Embeddings makes performing natural language and coding tasks easier.

Embeddings

13 of 23

13

Vector DB

Manually or using �an algorithm

Embedding model

VectorDB

Your knowledge base to the embeddings in the VectorDB

How to Create a Knowledge Base

Knowledge

Base

Chunk

(segment the knowledge)

Convert chunks to embeddings

Index the embeddings

to the VectorDB

14 of 23

Vectorization and Embedding

https://github.com/GaiaNet-AI/rag-tool

The embedding models are specially trained LLMs that turns input sentences into a vector representation, instead of generating completions. This allows the GaiaNet node to quickly identify contents in its knowledge base that are relevant to the current conversation.

08

What to build with Gaia

15 of 23

15

Node Demo

Scan the QR code and follow instructions on installing and interacting with your own node.

How to Create a Knowledge Base

16 of 23

16

Introduction to Gaia

How to launch an AI Agent

Demo of Gaia

Contributing to Open Source & Gaia

Table of Contents

17 of 23

17

Demo

Browse the agents

https://www.gaianet.ai/agents

Playing in Gaia

18 of 23

18

Introduction to Gaia

How to launch an AI Agent

Demo of Gaia

Contributing to Open Source & Gaia

Table of Contents

19 of 23

19

Contribute

CONTRIBUTE HERE

https://github.com/GaiaNet-AI/awesome-gaia

Playing in Gaia

20 of 23

20

Knowledge Builders🌱

Scan the QR code and join our telegram channel for help with any technical questions.

How to Create a Knowledge Base

21 of 23

Join the discussion

github.com/GaiaNet-AI

gaianet.ai

@Gaianet_AI

11

22 of 23

22

Any questions?

Playing in Gaia

@HarishKotra

23 of 23

hello@gaianet.ai

Thanks

Gaianet.ai

github.com/Gaia-AI

@Gaia_AI