Skip to main content
Loading navigation
Back to challenges

Build a Financial News Anomaly Monitoring Network with AutoGen

Sudden M&A rumors or deal collapses cause immediate volatility and potential market manipulation in listed financial stocks. Build a multi-agent conversational pipeline using Microsoft AutoGen to monitor streaming financial news and cross-reference order book anomalies. Detect market sentiment manipulation signals with 92% accuracy across breaking news streams.

CybersecurityHosted by Vera
Sign in free to start

What you are building

Construct an AutoGen multi-agent system to analyze breaking financial news signals and correlate them against real-time trading volatility.

How work is evaluated

Evaluation

Evaluate multi-agent system efficiency in classifying genuine news vs market manipulation signals.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

Detect Unverified Takeover Rumor

Assert manipulation_detected is true on unverified high-volume drops.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

Signal Precision

Precision of manipulation classification over benchmark news stream. • target: 0.92 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

  • Orchestrate multi-agent dialogue using AutoGen framework (UserProxyAgent, AssistantAgent)

  • Synthesize unstructured financial market rumors with order book order-flow dynamics

  • Identify coordinated market manipulation or deliberate misinformation campaigns

  • Emit structured risk telemetry alerts for compliance desk automated oversight

Resources

Dataset notes

Synthetic dataset of 1,000 paired financial market news stories and ticker volume time series.

Challenge setup

Evaluate multi-agent system efficiency in classifying genuine news vs market manipulation signals.

Challenge input

JSON containing news_headline, publication_timestamp, ticker, price_change_pct, and order_volume_spike.

AutoGen

Enables multi-turn conversational reasoning between specialized agent roles.

Evaluated output

JSON containing manipulation_detected (boolean), sentiment_impact ('HIGH', 'MEDIUM', 'LOW'), and explanation.

Evaluation checks
  • Assert manipulation_detected is true on unverified high-volume drops.
  • Precision of manipulation classification over benchmark news stream. • target: 0.92 • range: 0-1
Configured evaluation evidence
  • Benchmark: MarketFinEval
  • Signal Precision target: 0.92
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • AutoGen · Required
  • Microsoft AutoGen · Optional

Evaluation contract

  • Detect Unverified Takeover Rumor · Weight 1
  • Signal Precision · Weight 1

Metrics

  • Signal Precision · Target 0.92

Benchmarks

  • MarketFinEval
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing news_headline, publication_timestamp, ticker, price_change_pct, and order_volume_spike.

    Output: JSON containing manipulation_detected (boolean), sentiment_impact ('HIGH', 'MEDIUM', 'LOW'), and explanation.

    Execution harness: Python

Evaluation tools

  • AutoGen · Multi-Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a Financial News Anomaly Monitoring Network with AutoGen