Spicy AI Collaboration with Storacha - Apr 14th - TripMate - Encode AI Hackathon

[Dhruv-Varshney-developer]

Spicy AI Collaboration with Storacha - Apr 14th - TripMate - Encode AI Hackathon

Links

• TripMate Planner - Agent 1
• TripMate Finder- Agent 2
• TripMate Share- Agent 3
• Technical Whitepaper
• Tripmate Overview
• Presentation Link

Contributors

@Dhruv-Varshney-developer @Nkovaturient

Accomplished Tasks

• Built three specialized AI agents that collaborate through Storacha storage
• Implemented Agent-1 (TripMate Planner) with Telegram interface and Storacha integration
• Created Agent-2 (TripMate Finder) with WebSocket communication to process travel requests
• Developed Agent-3 (TripMate Share) for retrieving and sharing stored trip data
• Successfully deployed Agent-1 and 2 on Railway for production use
• Created comprehensive technical documentation and whitepaper

Update

• We've built TripMate, a multi-agent AI travel planning system that leverages Storacha's decentralized storage
• The system consists of three specialized agents working together:
  • TripMate Planner (Agent-1): Telegram bot that collects user preferences and orchestrates the planning process
  • TripMate Finder (Agent-2): Python service that processes user preferences to search for travel options
  • TripMate Share (Agent-3): Enables viewing stored trip details and adding suggestions
• What went well:
  • Integration of Storacha with ElizaOS was smooth once we understood the flow
  • The multi-agent architecture effectively demonstrates collaborative AI
  • WebSocket communication between agents works efficiently
• What was difficult:
  • Memory constraints when deploying Agent-1 and Agent-3 on Railway
  • Couldn't integrate Agent-2 with Storacha directly due to Python client limitations
  • Managing environment variables across multiple services was challenging

Demo Videos

https://www.veed.io/view/e90b293f-0748-4790-9267-b9aa90c7b180?panel=share

Feedback

• The Storacha team was incredibly supportive throughout our development process
• Documentation was comprehensive and easy to follow
• We would have liked to integrate our Python-based Agent-2 directly with Storacha, but the Python client is still under development
• The overall experience with the technology was excellent, and we're excited to see how the ecosystem continues to evolve

Next Tasks

• Improve error handling across all agents
• Reduce memory footprint of Agent-1 and Agent-3 to optimize deployment
• Add a new booking agent with cryptocurrency payment options
• Integrate OpenAI Search Kit alongside SERP API for improved real-time data
• Expand to platforms beyond Telegram
• We're interested in continuing development on this project and participating in the microgrants program

[Nkovaturient]

Great collaboration⚡🌟💯

• From sketching rough outlines to bringing every line of code to life, this journey was nothing short of a rollercoaster — filled with perplexing moments, bugs that tested our patience, and pivots that made us rethink and rebuild.

• Through high tides of breakthroughs and low tides of debugging marathons, every challenge pushed us closer to a smarter solution. Each modification wasn’t just a fix — it was a raw level-up!

• Building TripMate, our Multi-Agent Planner, is not just a project — it's a milestone. A true victory in motion. 🚀