Inspiration
In today’s job application ecosystem, candidates unknowingly share sensitive personal data across multiple platforms—with little visibility or control. While recruiters can easily access candidate information, users are left in the dark, unable to track who is accessing their data, how often, or for what purpose. This imbalance creates a critical data trust gap between job seekers and recruiters. Built for AI Automate 2026, we designed DataStalker AI to close this gap—an autonomous system that can observe, decide, and act on data access in real time.
What it does
DataStalker AI is a lightweight autonomous AI agent that monitors, analyzes, and controls how recruiters access user data.
The system:
Tracks every data request made by recruiters
Calculates a dynamic Risk Score based on behavioral patterns
Automatically decides whether to Allow, Deny, or Mask data
Displays real-time insights through an interactive dashboard
Instead of passive data exposure, users gain active, intelligent protection.
How we built it
We built DataStalker AI using a simple yet powerful architecture:
Frontend: HTML, CSS (interactive dashboard)
Backend: Python (Flask)
Database: SQLite (stores requests, decisions, and scores)
AI Logic: Lightweight autonomous AI agent powered by a rule-based decision engine
Risk Score Logic: Risk = (Frequency × 2) + (Denied × 3) + (OddTime × 2) + (Unknown × 5)
Decision Engine: High Risk → Deny Medium Risk → Mask Data Low Risk → Allow This enables fast, transparent, and explainable decisions without relying on heavy machine learning models.
Key Features
- Autonomous AI Decision Engine
Automatically evaluates and responds to data access requests
- Live Attack Simulation Simulates suspicious recruiter behavior to demonstrate real-time AI defense
- Stalker Leaderboard Ranks companies based on risk score and access patterns
- Explainable AI Clearly shows why each decision was made
- Data Masking Protects sensitive information by partially hiding it instead of fully blocking access
Challenges we ran into
Making the system feel intelligent while keeping it fast and lightweight
Balancing automation with user control
Simulating realistic recruiter behavior without real-world datasets
Creating an engaging demo under tight time constraints
Accomplishments that we're proud of
Built a fully functional autonomous decision-making system
Designed a practical and explainable risk scoring model
Developed a compelling, interactive demo experience
Addressed a real-world problem in data privacy and transparency
What we learned
Simplicity, when designed well, can rival complex AI systems
Data transparency is a major gap in modern digital platforms
Strong storytelling significantly improves how projects are perceived
Prioritizing core functionality is key in time-bound builds
What's next for DataStalker AI: Autonomous Data Guardian
Integration with real-world platforms (LinkedIn, job portals)
Adding machine learning for advanced behavioral prediction
Real-time alerts for suspicious activity
Expansion into domains like fintech, healthcare, and cybersecurity.
