Signal Sync

Agentic AI-powered BGV verification system that analyzes companies, detects financial risks, and identifies potential scams to help retail investors make informed decisions.

Stack

Python CrewAI Streamlit yFinance OpenAI API
Signal Sync

Overview

Signal Sync is an AI-driven background verification system designed for retail investors. It uses a multi-agent architecture (CrewAI) to analyze companies across multiple dimensions, including management credibility, financial integrity, and market behavior, and generates a structured risk report with actionable insights.

The system combines real-time stock data, document analysis, and web intelligence to provide a comprehensive investment risk assessment.

Key Features

  • Multi-agent AI system for automated company verification
  • Company overview analysis including business model and market presence
  • Management background checks to identify red flags or controversies
  • Financial irregularity detection (debt patterns, anomalies, audit concerns)
  • Scam detection using stock price and volume analysis
  • Structured JSON output with risk scores and final verdict
  • Interactive Streamlit dashboard for input and visualization

Tech Stack

  • Frontend: Streamlit (interactive dashboard)
  • Backend: Python with CrewAI
  • Data: yFinance for stock data
  • AI: OpenAI API for analysis
  • Validation and Schema: Pydantic models
  • Configuration: YAML-based agent and task system

Architecture

Signal Sync follows a modular multi-agent architecture:

  • Agents Layer: Defined in YAML (research, financial, scam detection agents)
  • Task Pipeline: Sequential execution of verification steps
  • Data Layer: Stock data and document inputs
  • Processing Layer: AI-driven analysis and anomaly detection
  • Output Layer: Structured JSON report with scores and findings

Output

The system generates:

  • Trustworthiness Score
  • Financial Integrity Score
  • Management Risk Score
  • Market Manipulation Risk Score