New Platform 'Tremor.live' to Revolutionize Prediction Market Analysis

A former Instacart engineer has unveiled a new artificial intelligence-powered platform designed to detect significant fluctuations within prediction markets. This innovative tool, named Tremor.live, monitors trading activity on platforms such as Polymarket, aiming to pinpoint sudden price shifts that could signal the presence of insider information or emerging trends before they become widely known. By integrating with social media and utilizing AI for analysis, Tremor.live offers a dynamic new approach to understanding market movements and potentially gaining an edge in forecasting future events.
Nikshep Saravanan, the creator behind Tremor.live, developed the platform using Claude Code and deployed it via Vercel. His primary motivation was to create a system capable of identifying unusual volatility in prediction markets. These markets allow participants to wager on the outcomes of various future events, ranging from political elections to technological advancements. Saravanan explains that unexplained jumps in the probability of an event, such as a ceasefire in a conflict, could indicate that individuals with privileged information are influencing the market without any corresponding public news. This tool seeks to highlight such anomalies.
Currently, Tremor.live actively monitors approximately 500 prediction markets on Polymarket. It tracks price changes in real time and assigns an intensity score from 0 to 10 to each movement. Scores above 7.5 are categorized as 'Extreme,' indicating major market shifts, while scores between 5 and 7.5 are 'High.' 'Moderate' movements fall between 2.5 and 5, and 'Low' movements, representing minor fluctuations, are between 0 and 2.5.
Beyond just tracking price changes, Tremor.live incorporates an AI-powered analysis of social media. It integrates with platforms like X (formerly Twitter) through Grok's API to scan for relevant discussions that might explain market movements. This feature allows the platform to provide users with summaries, helping them understand whether market shifts are driven by publicly available information or if the lack of public discussion suggests more opaque, potentially insider-driven activity.
An example of Tremor.live's utility was observed when it flagged a prediction market concerning Tesla's robotaxi launch in California by 2025. The market showed an 'Extreme' intensity rating of 7.7. Tremor's AI-generated summary suggested that traders were anticipating a delay or cancellation, citing insights gleaned from insider posts on X pointing to unresolved technical or regulatory challenges. This illustrates how the platform can provide timely, actionable intelligence.
Still in its early stages of development, Tremor.live is continually being refined with new features. Saravanan has also made the project's code available on GitHub, encouraging broader exploration and collaboration within the tech community.
Tremor.live provides a novel mechanism for participants in prediction markets to identify unusual trading patterns and potentially gain insights into future events. By combining real-time market monitoring with AI-driven social media analysis, the platform aims to empower users to differentiate between publicly influenced trends and those potentially shaped by non-public information, offering a distinct advantage in the dynamic world of predictive forecasting.