Skip to main content

Data Barista: Semantic Matchmaking Agent Demo

Introduction [0:00-2:00]​

  • Eliza's Builder Demos featuring Amir from Data Barista
  • Data Barista is a "super connector" Eliza OS agent for semantic matchmaking
  • Goal is connecting people based on mutual interests and needs

About the Creator [2:00-6:00]​

  • Amir Alkhani, an engineer who switched from chip design to AI agents in crypto
  • Started with Eliza OS a few months ago, found it modular and easy to use
  • Previously worked on data DAOs (Netflix, Twitter) to give data ownership to users

The Problem Data Barista Solves [6:00-9:00]​

  • 50% of romantic matchmaking happens online through algorithms
  • Technical networking still relies on inefficient methods:
    • Conferences with rushed conversations
    • Digital networking with anonymous users on Telegram/Discord
    • Difficulty finding relevant connections among many introductions
    • Lack of verification for people's claimed backgrounds

Live Product Demo [9:00-16:00]​

  • Amir demonstrated creating a new profile for a fitness agent project
  • Agent asked targeted questions about the project details
  • Example: Agent learned about a fitness app that judges food pictures post-workout
  • The agent then connected Amir with a community builder as a match

How Data Barista Works [11:00-14:30]​

  • Extracts knowledge triplets from conversations (goals, challenges, connections)
  • Separates data into public (anonymized on OriginTrade) and private pieces
  • Private data stored in edge node only accessible to the agent
  • Uses semantic matchmaking through:
    1. Initial query for potential matches
    2. Vector database collection
    3. Ranking system to find top 10 matches
    4. Posts introduction to the best match

Future Vision [14:30-17:30]​

  • Building a grassroots data DAO with:
    • User privacy and data ownership through NFTs
    • Token rewards for sharing data with the agent
    • Users can sell tokens on the market
    • Creates a non-speculative token economy backed by data
  • Modular design to work on any client for various matchmaking purposes
  • Transparent governance of data rules and consent mechanisms
  • Currently seeking builders who share this vision

Call to Action [17:30-18:20]​

  • Invitation to test the agent on Telegram, Twitter, and the Data Barista web interface
  • Looking for collaborators to help build grassroots data DAOs

The demo showed how Data Barista can efficiently collect relevant information from users and provide meaningful connections based on complementary interests and needs, all while maintaining data privacy.