Real-Time Intrusion Detection Platform

Overview

Built a production-grade system to detect and visualize intrusion events using live data from fibre optic sensing hardware. The platform processes high-frequency sensor input, detects anomalies, and provides real-time alerts along with visual context via CCTV streams and interactive dashboards.

This project combined hardware integration, distributed systems architecture, and real-time data processing to create a comprehensive security solution for critical infrastructure monitoring.

Spring Boot PostgreSQL InfluxDB AWS Raspberry Pi React RTSP

Key Highlights

Hybrid Architecture
Edge processing on Raspberry Pi + cloud-based services on AWS for scalable, resilient monitoring

Real-Time Pipeline
Ingested and analyzed continuous high-frequency sensor data using InfluxDB for time-series storage

Live Monitoring
Map-based interactive UI with intrusion visualization and real-time system metrics

Video Integration
RTSP-based CCTV streaming embedded directly into the web interface for context-aware alerts

Smart Alerting
Multi-channel notifications (WhatsApp, Telegram, Email, ntfy) for instant incident response

Security First
Authentication, IP-restricted deployments, and secure credential handling for sensitive environments

Technical Approach

Designed and built a production-ready real-time intrusion detection system integrating fibre optic sensing hardware with cloud and edge infrastructure
Developed an event-driven backend using Spring Boot to ingest high-frequency sensor data from fibre optic arrays
Built dual-database strategy: PostgreSQL for transactional data (events, users, alerts) and InfluxDB for time-series sensor metrics
Implemented edge + cloud architecture with on-site Raspberry Pi nodes running anomaly detection algorithms and AWS-based notification services
Engineered a multi-channel alerting system (WhatsApp, Telegram, Email, ntfy) with intelligent throttling and escalation logic
Built live CCTV streaming pipeline (RTSP) and seamlessly integrated video feeds into a React-based monitoring dashboard
Created an interactive map-based UI to visualize intrusion events, exact locations, and system status in real time
Designed and implemented comprehensive user and notification management systems with role-based access control and secure authentication
Collaborated directly with hardware engineers to align software architecture with fibre optic sensing modules and constraints
Led a team of developers and drove end-to-end delivery from requirements gathering to on-site deployment and training

My Role & Impact

Leadership & Ownership

Led system architecture design and technology decisions for a greenfield project
Managed a cross-functional team of backend and frontend developers
Acted as the bridge between hardware engineers and software implementation
Delivered the entire system end-to-end, including on-site deployment, configuration, and stakeholder training

Technical Contributions

Approximately 60% hands-on code contribution across backend, infrastructure, and deployment
Architected the event-driven data pipeline for real-time sensor processing
Implemented edge computing strategy for low-latency anomaly detection on Raspberry Pi
Designed the multi-layer alerting system with reliability and escalation logic
Mentored team members on distributed systems concepts and cloud deployment practices

What This Demonstrates

Problem-to-Product Thinking: Took a hardware constraint (fibre optic sensing) and built a complete end-to-end solution
Rare Skill Combination: Hardware integration + real-time systems + distributed architecture is uncommon and valuable
Ownership Mindset: Led from requirements to deployment; didn't just code features but owned outcomes
System Design Thinking: Made thoughtful tradeoffs (edge vs. cloud, time-series vs. relational storage, multiple alert channels)
Team Leadership: Managed developers, collaborated with hardware engineers, communicated with stakeholders
Production Maturity: Built security, monitoring, and operational readiness into the system from the start