Replay Ai

Provide a short tagline for the project. You can change this later.

Comment

About the Project Inspiration

The inspiration for RelayAI came from the desire to build an AI-powered system that could not only execute complex tasks for users but also provide immutable proof that these tasks were completed. With the growing use of artificial intelligence in critical systems, it's important to add an extra layer of trust and transparency. By integrating Solana blockchain, we can provide verifiable proofs for every action the AI performs, ensuring that tasks are not only automated but also auditable and tamper-proof. The goal was to create a system that merges AI with blockchain to bring both efficiency and transparency to the table.

What It Does

RelayAI is an AI-powered assistant that performs various tasks based on user input, such as:

Searching for cheapest flights to a destination.

Performing research tasks.

Offering personalized recommendations.

What makes RelayAI unique is its integration with the Solana blockchain. Every task the AI completes is recorded as a transaction on Solana, providing a proof of execution that is immutable and verifiable. This ensures accountability and transparency for users interacting with the AI.

How We Built It

We used the following technologies to build RelayAI:

Google Gemini API: Used to process natural language queries, generate responses, and break down tasks.

Node.js and Express: Built the backend API that connects the frontend with the AI and Solana blockchain.

Solana Blockchain: Used to store proof of task execution. Every task is verified with a transaction signature.

DigitalOcean: Deployed the backend API on DigitalOcean Droplets for reliable and scalable hosting.

ElevenLabs Text-to-Speech (TTS): Integrated to provide spoken responses for users, enhancing interactivity.

We also built a simple web interface where users can type their queries and receive results, including the Solana transaction signature as proof of the completed task.

Challenges We Ran Into

Blockchain Integration: Setting up Solana and making sure the proof was correctly recorded on the blockchain was challenging. Solana’s rate limits for airdrops caused delays in testing, and initially, I had trouble understanding how to generate and use the correct Solana keypair.

AI Response Structuring: Getting Google Gemini to generate coherent and structured responses, especially when breaking down complex tasks, required fine-tuning of the prompts and proper handling of edge cases.

Hosting and Deployment: Deploying the app on DigitalOcean and ensuring that all environment variables (API keys, Solana key, etc.) were properly configured was a challenge. Ensuring the app could scale while maintaining security was also a key hurdle.

Accomplishments That We're Proud Of

Immutable Proofs on Solana: We successfully integrated Solana blockchain to store immutable proofs of AI actions, making the entire system transparent and auditable.

Seamless Integration of AI and Blockchain: Combining AI task execution with blockchain verification is not something commonly done, and we were able to create a proof-of-concept that combines both efficiently.

Interactive AI Assistant: The integration of ElevenLabs Text-to-Speech adds a unique, interactive component to the AI, making it more engaging for users.

What We Learned

Blockchain Fundamentals: I gained a deeper understanding of Solana blockchain, including how to interact with it programmatically and use it to store proof.

AI Response Generation: Learning how to craft prompts for the Google Gemini API was crucial to ensuring the AI generated relevant and structured responses for the tasks at hand.

Full-Stack Development: I learned how to deploy a Node.js application to a cloud provider like DigitalOcean, setting up continuous hosting and scaling options.

What's Next for RelayAI

Integration with More APIs: Adding more third-party APIs (such as flight booking systems, weather data, or e-commerce APIs) to make RelayAI even more powerful and capable of handling more diverse tasks.

Mobile App: We plan to build a React Native mobile app so users can interact with RelayAI on their smartphones and receive notifications or spoken responses directly.

Solana Smart Contracts: Exploring the use of Solana smart contracts (Programs) to automate certain tasks and actions, making the system more autonomous.

User Authentication: Adding user login and profile features to personalize the AI's responses and provide a better user experience.

Multi-Language Support: Adding multi-language support to reach a global audience and make RelayAI more inclusive.

Built With

Share this project:

Updates