|
Core Focus: Systems Architecture, Network Security, and Low-Level Engineering. |
I specialize in optimizing system performance, securing networks, and applying intelligence to critical infrastructure.
🚦 The Route Cause (Smart City Infrastructure)
- Problem: Static traffic signals fail to handle surges or emergency vehicles.
- Solution: Engineered a 3-Tier Intelligent Architecture acting as Eyes (Perception), Brain (Decision), and Supervisor (Guardian).
- Tech: Utilized YOLOv8 for density detection and a Q-Learning Agent for signal timing.
- Key Feature: The "Guardian Layer" enforces safety rules, granting immediate preemption for emergency vehicles and preventing starvation.
⚡ AI-PoT Scheduler (Distributed Systems)
- Goal: Reducing P99 (Tail) Latency in heterogeneous clusters where classical Power-of-Two choices fail.
- Implementation: Built a simulation using MPI (OpenMPI) to model straggler nodes and heavy-tailed "elephant" workloads.
- Innovation: Integrated an O(1) inference Linear Regression model to predict task costs, outperforming classic schedulers in high-load scenarios.
📦 Process Migration Engine (Load Balancing)
- Metric: Orchestrated proactive container migrations using CRIU and Docker.
- Result: Improved load distribution by 25% with predictive ML models.
| Domain | Tools & Technologies |
|---|---|
| Network & Security | Burp Suite, Socket Programming, TCP/IP, OSI Layers, Security Fundamentals |
| Systems Programming | C, C++, MPI (OpenMPI), Linux/Unix Shell, CRIU |
| AI & Simulation | YOLOv8, Q-Learning (RL), Scikit-learn, Pandas |
| Backend & DevOps | Python, FastAPI, Docker, Vagrant, WebSockets |
- Focus: System Latency, Network Resilience, Gridlock Prevention.
- Architecture Style: Event-Driven, Master-Worker, Multi-Tiered AI.
- LinkedIn: linkedin.com/in/bind-pratap-singh
- Email: bs563@snu.edu.in