0xKnowledge

• 

  0xKnowledge - AI assistants for every human

• 

  Slide 2 - Problem statement

• 

  Slide 3 - Real costs and problem sizing

• 

  Slide 4 - Market opportunity

• 

  Slide 5 - 0xKnowledge

• 

  Slide 6 - Product overview

• 

  Slide 7 - Bedrock

• 

  Slide 8 - Bedrock AI

• 

  Slide 9 - Ember AI Agent Builder

Inspiration

As blockchain enthusiasts, developers and community builders, we've experienced firsthand the challenges of navigating fragmented and outdated documentation in the Web3 space. We envisioned a solution that not only streamlines documentation with Web3-native components, but also democratizes AI agent creation, making cutting-edge technology accessible to all.

What it does

0xKnowledge offers a Web3-native, AI-augmented documentation platform that revolutionizes how blockchain projects manage and share knowledge. This dev tool includes:

• No-code editor to edit knowledge bases
• Publish knowledge bases as developer documentation sites
• AI-powered search and a help agent that answers user questions
• Real-time blockchain data panel that shows smart contract data

How we built it

We leveraged a stack of cutting-edge technologies to build a Web 2.5 platform. This includes Next.js on the frontend and Supabase and Postgresql on the backend for the web application. We also used Pinecone for our vector database, and LlamaIndex for AI orchestration. Finally, we used Alchemy and Dune to retrieve onchain data from smart contracts.

Challenges we ran into

• Optimizing AI response times for user questions on the documentation sites
• Deploying the documentation sites and updating them (with lower delays)
• Balancing user and reader friendliness with the features of an AI chat (not overcrowding user interfaces)
• Integrating blockchain data sources on documentation sites

Accomplishments that we're proud of

• Developing a documentation platform that deploys live documentation sites with organizations and a multi-tenant structure
• Connecting our knowledge bases and enabling Retrieval Augmented Generation on both the docs and internal knowledge bases
• Connecting our chat with on-chain data to respond with quantitative data based on natural language prompts
• Connecting our docs sites with on-chain data to create a playground for end-users

What we learned

• The importance of reader and writer experience in documentation
• Techniques for optimizing AI performance to answer user prompts in a contextualized way
• The potential of combining AI with blockchain for knowledge management

What's next for 0xKnowledge

• Refining on the onchain data integrations to build a better user interface for the data playground/panel
• Expanding blockchain integrations to support more networks like other EVM chains, Solana, Move chains, and ICP smart contract data
• Implementing advanced analytics for documentation usage and AI agent performance
• Creating an interopable decentralized application for users to deploy knowledge assets across multiple chains that help with AI-related operations on-chain using Coinweb

Built With

• llamaindex
• next.js
• pinecone
• react.js
• supabase