Financial Data · Risk · Applied AI

Market Whisperer — Event-Driven Portfolio Risk Prototype

Built a hackathon prototype combining event, sentiment, prediction-market, and financial-news signals to explore short-term changes in portfolio volatility and risk.

Geopolitical risk map produced by the Market Whisperer prototype
Prototype risk view
Role
HackEurope 2026 project
Date
2026
Context
Hackathon prototype
Recognition
Top 3 — Data Analysis

Technologies

PythonMarket dataSentiment analysisNews & prediction-market dataRisk analysis

Context

Context

Market Whisperer is a hackathon prototype, built at HackEurope 2026, that combines event, sentiment, prediction-market, and financial-news signals to explore how short-term portfolio volatility and risk might shift. It earned a Top 3 placement in Data Analysis. It is deliberately framed as an exploratory prototype — not a production trading system.

Problem

Problem

Risk-relevant information is scattered across heterogeneous sources — market data, news, sentiment, and prediction markets — with very different structures and reliability.

Sentiment and event signals are noisy and easy to over-interpret, especially under hackathon time pressure.

Building something demonstrable in a short window required ruthless scoping toward an exploratory prototype.

Scope

System scope

  • Market and financial-news data ingestion

  • Sentiment and event signals

  • Prediction-market information

  • Exploratory portfolio-risk view

Methods

Approach & methods

  • Ingested market and financial-news data alongside sentiment and prediction-market signals.

  • Combined these heterogeneous signals to explore short-term changes in volatility and portfolio risk.

  • Prioritized rapid product development to reach a working, demonstrable prototype within the event.

Contributions

Contributions

  • Contributed to a team project at HackEurope 2026 that placed Top 3 in Data Analysis.

  • Focused on combining event, sentiment, and prediction-market signals into an exploratory risk view.

Results

Results

Recognition

Top 3

Data Analysis, HackEurope 2026

Format

Hackathon

Exploratory prototype

Process

Technical process

  1. 01Market & news data
  2. 02Sentiment & events
  3. 03Signal fusion
  4. 04Risk view

Limitations

Limitations

This is an exploratory hackathon prototype: it makes no claim of trading performance, alpha, Sharpe ratio, or backtested returns.

Signals were not validated for production use and were not deployed.

Hackathon time constraints limited the depth of evaluation and data coverage.

Report

Report & resources

Risk engine write-up (PDF)
Financial Data · Risk · Applied AIHackEurope 2026 project