Insights — AI-Powered Security Intelligence Platform

Turn Queries into Clarity. Analyze logs, detect anomalies, and secure your systems with a context-aware AI engine designed for security professionals.

VISIT: https://insights-analytics.vercel.app/

Overview

Insights is a full-stack AI Security Operations Center (SOC) platform that transforms raw security logs into actionable intelligence. It combines:

• Natural Language Querying — ask questions about your logs in plain English
• Automated Threat Classification — Regex + LLM hybrid pipeline classifies every log entry
• Automated SOC Threat Sweep — batch-processes profiles to surface hidden threats
• SQL Investigation Console — direct forensic database access for advanced analysts
• Digital Twin Network Simulation — Cytoscape.js visualization of attack propagation

---

Features

Feature	Description
📂 Log Ingestion	Upload security CSV logs per project; AI classifies them instantly
🤖 AI Classification	Regex fast-path + Llama 3.3 (Groq) for unmatched entries
💬 Conversational Chat	NL→SQL engine answers security questions in natural language
🔍 Threat Sweep	Automated batch SOC analysis across all suspicious IP profiles
🗄️ SQL Console	Write and execute raw SQL directly against the project database
🌐 Digital Twin	Interactive network graph simulating attack vectors
📊 Multi-Project	Isolated per-project SQLite databases with JWT-authenticated access

---

Tech Stack

Frontend

• Next.js 16 — React framework with standalone output
• Tailwind CSS v4 — Utility-first styling
• Cytoscape.js — Network graph visualization
• Lucide React — Icon library

Backend

• Flask 3 — Python web framework
• PyJWT — JWT authentication
• SQLite — Per-project log databases

AI Engine

• Groq — Ultra-fast LLM inference (Llama 3.3-70B)
• Pandas — Log processing and schema normalization
• scikit-learn — ML utilities

---

Project Structure

Hackup/
├── Backend/                  # AI engine (classification + threat sweep)
│   ├── main.py               # Core: NL2SQL, threat sweep, DB helpers
│   ├── classification_log.py # Log classification pipeline
│   ├── llm_prompting.py      # Groq LLM prompts
│   ├── requirements.txt      # AI engine Python deps
│   └── data/                 # Per-project SQLite databases (gitignored)
│
├── frontend/
│   ├── backend/              # Flask REST API
│   │   ├── app.py            # All API routes
│   │   └── requirements.txt  # Flask Python deps
│   │
│   └── frontend/             # Next.js app
│       ├── src/
│       │   ├── app/          # Next.js pages (App Router)
│       │   ├── components/   # React components
│       │   └── context/      # DashboardContext (global state)
│       ├── public/           # Static assets
│       └── next.config.ts    # API proxy rewrites
│
├── Dockerfile                # Multi-stage production Docker build
├── start.sh                  # Container startup script
├── .dockerignore
├── .gitignore
├── DEPLOY.md                 # Google Cloud Run deployment guide
└── .env                      # Local secrets (gitignored)

---

Running Locally

1. Start the Flask backend

cd frontend/backend
pip install -r requirements.txt

# Also install AI engine deps
pip install -r ../../Backend/requirements.txt

python app.py
# Runs on http://localhost:8000

2. Start the Next.js frontend

cd frontend/frontend
npm install
npm run dev
# Runs on http://localhost:3000

Open http://localhost:3000 in your browser.

---

API Reference

Method	Endpoint	Description
POST	/api/login	Login, returns JWT cookie
GET	/api/projects	List all projects for current user
POST	/api/project	Create a new project
POST	/api/project/upload	Upload & classify a CSV log file
POST	/api/project/:id/query	Natural language log query
GET	/api/project/:id/threat-sweep	Run full SOC threat sweep
POST	/api/sql	Execute raw SQL against project DB
DELETE	/api/project/:id	Delete a project and its database

---

Team

Name	Email
Mitansh Jadhav	mitansh.jadhav2007@gmail.com
Om Korade	omkorade23@gmail.com

---

© 2026 Insights Intelligence. Built for security professionals.