Search projects, hackathons, sponsors and tech

    ← All projects

    Rabble Reimagine: Innovate & Elevate

    Level up your crypto game with Rabble's latest update! Our new AI-powered features make it easier than ever to stay informed, make informed decisions, and maximize your crypto potential.

    About

    The problem it solves The cryptocurrency landscape is inundated with information, making it challenging for traders to navigate misinformation, stay updated on market trends, and access actionable insights. Rabble’s current features provide transactional benefits, but users face the following pain points: 1. Misinformation in Telegram Crypto Channels Challenge: Telegram channels are rife with unreliable predictions, making it hard for users to trust investment advice. This often leads to poor decision-making and reduced confidence. Solution for Rabble: Rabble’s NLP-based prediction validation system analyzes Telegram channels, scoring their prediction accuracy by using historical data and market outcomes. Verified scores provide users with trustworthy insights, making Rabble a credible source for crypto decisions. 2. Overwhelming Crypto Data Challenge: Users are overwhelmed by the sheer volume of daily crypto news, trends, and market changes. Beginners especially struggle to focus on what matters most. Solution for Rabble: A ML-powered curated feed simplifies this complexity by providing a personalized daily feed of the most relevant crypto updates. Rabble users can stay informed without wasting time filtering through unnecessary information. 3. Fragmented Token Analytics Challenge: Traders often need to visit multiple platforms for token-specific data like transactions, price trends, and market activity, slowing decision-making. Solution for Rabble: The Token Activity Dashboard consolidates token-specific metrics, offering a centralized, interactive interface with transaction history, price alerts, and key market insights. This feature enhances user decision-making and trading efficiency. Impact on Rabble By solving these challenges, Rabble evolves into a comprehensive platform for crypto enthusiasts: Builds User Trust: Reliable insights validated by NLP technology enhance decision-making confidence. Simplifies Crypto Engagement: Personalized feeds and dashboards improve accessior

    How it's made

    Challenges we ran into 1. Handling Misinformation in Telegram Channels Challenge: Extracting and analyzing predictions from Telegram channels was difficult due to the unstructured nature of the data. Messages often contained inconsistent formats, abbreviations, and noise like emojis or links. Solution: We implemented advanced Natural Language Processing (NLP) techniques for data preprocessing, including text normalization and sentiment analysis. Using a combination of models like BERT and custom rule-based filters, we improved the accuracy of data extraction and analysis. 2. Building a Real-Time Curated Feed Challenge: Integrating multiple data sources (news, social media, on-chain analytics) while ensuring relevance and timeliness was complex. The feed had to cater to diverse user preferences without overwhelming them. Solution: We designed a hybrid ML model combining content-based and collaborative filtering to personalize the feed. Incremental learning techniques ensured the feed adapted to new data in real time, and A/B testing helped refine user engagement. 3. Scaling the Token Activity Dashboard Challenge: Fetching and visualizing real-time transaction data and price alerts for multiple tokens created performance bottlenecks. Solution: We optimized backend APIs using caching and implemented WebSocket connections for real-time updates. An interactive dashboard with lightweight visualizations reduced client-side load. 4. Ensuring User Trust Challenge: Users were skeptical about the accuracy of the prediction validation and feed recommendations. Solution: We added transparent scoring metrics and visualized channel performance over time, enabling users to evaluate the reliability of insights themselves. Continuous user feedback helped refine the models further.