Back to projects
Rohit Bamhande

phaiAI Fellow

Rohit Bamhande

phaiAI Fellow

Every packet has a story. LLMs help us hear it.

Team GuardGPT built a real-time intrusion detection system that classifies network anomalies with Isolation Forest, auto-maps alerts to MITRE ATT&CK, and escalates critical incidents to a Groq-hosted LLaMA 3.1 for natural-language triage.

In their words

My experience with phaiAI has been excellent. The internship provided a practical learning environment where I gained hands-on experience in AI concepts while working on real-world tasks. The mentors were supportive, approachable, and always ready to guide us whenever needed. The structured learning approach, practical assignments, and collaborative environment helped me improve both my technical and problem-solving skills. I also developed a better understanding of teamwork, project execution, and industry expectations.
Rohit Bamhande, Symbiosis Skills and Professional University

6

Tools & Frameworks

3

Models

5

AI Concepts

80%

Avg. Proficiency

AI Proficiency

76%

Network Security

80%

Anomaly Detection

81%

LLM Integration

84%

Real-time Systems

Tools & Frameworks

ScapyWiresharkPythonStreamlitFastAPIGroq API

Models

Isolation ForestLLaMA 3.1Random Forest

Concepts

NIDSMITRE ATT&CKPacket AnalysisThreat IntelligenceSOC Automation
Anomaly-Based Intrusion Detection System

Project · Cybersecurity

Anomaly-Based Intrusion Detection System

Team GuardGPT

A real-time, AI-enhanced intrusion detection system that captures network packets, classifies anomalies with an Isolation Forest model, and auto-tags every alert with MITRE ATT&CK tactics. High and critical alerts are escalated to the Groq LLaMA 3.1 API for natural-language threat analysis and remediation, all surfaced on a live monitoring dashboard.

Isolation ForestMITRE ATT&CKGroq LLaMA 3.1Real-time NIDS

Built alongside