PredictChain: AI-Powered Stock Predictions

About

The problem it solves PredictChain aims to integrate AI-based stock market predictions with blockchain to provide decentralized and tamper-proof access to trading insights. This can help traders and DeFi platforms make informed financial decisions while ensuring data integrity and transparency through smart contracts. Though incomplete, the envisioned system would: Use Chainlink oracles to securely bring AI predictions on-chain. Enable transparent decision-making with decentralized data validation. Enhance the DeFi ecosystem with reliable predictive analytics Additional Features Although the project remains incomplete, I explored the following areas during the hackathon: Experimented with Chainlink for setting up a decentralized oracle to connect AI predictions to the blockchain. Tested AI-generated stock market trends using TensorFlow to create a predictions dataset (predictions.json). Started drafting a Solidity-based smart contract to log predictions on-chain for future scalability. Gained hands-on experience with Docker, Chainlink, and smart contract workflows.

How it's made

Challenges we ran into Time Constraints: With limited time during the hackathon, integrating all components of the project proved challenging. Balancing between AI predictions, Chainlink setup, and smart contract development was overwhelming. Chainlink Node Setup: Setting up the Chainlink node and understanding its architecture took longer than anticipated. Issues with Docker configurations and missing files further delayed progress. Integration Complexity: Combining off-chain AI predictions with on-chain smart contracts required deep knowledge of oracles and adapters, which was challenging to achieve within the timeframe. Learning Curve: Using sponsored tools like Chainlink for the first time added to the learning curve, as I had to familiarize myself with their documentation and development workflows.