Hi, I'm _
IIIT Bhopal CS '27 operating at the intersection of systems engineering, offensive security, and air-gapped AI infrastructure.
Specializing in High-Performance C++ Engines · Vector Databases · Low-Level Security
01. Applied Research
Volatile Memory Forensics Engine
GitHub Repository ↗Context: A forensic security tool for detecting fileless malware and process-injection attacks by parsing and analyzing raw memory dumps — evading standard disk-based defenses.
Parses process memory space directly using Python + psutil to identify injected shellcode, hollowed processes, and anomalous memory regions that standard AV scanners miss entirely.
Levelled Homomorphic Encryption for O(1) Authentication
PQC RESEARCHContext: A post quantum batch authentication architecture utilizing the Brakerski Fan Vercauteren (BFV) homomorphic encryption scheme for IoT healthcare networks. Solves the O(N) verification bottleneck.
02. Experience
Software & Systems Security Intern
@ NIT Bhopal | Dec 2024 – Feb 2025
- Engineered async Python pipelines to parse high-volume system logs — replaced manual monitoring with automated threat anomaly detection
- Built dynamic log extractors for hybrid Linux/Windows environments, converting raw event streams into structured telemetry
- Mentored 250+ students in algorithmic threat modeling and secure system architecture
Backend & Security Engineering Intern
@ TCR Security | Oct 2024 – Dec 2024
- Architected Python async data ingestion engine for threat intel feeds — eliminated 20+ hours of manual processing/month
- Debugged web architectures via raw HTTP traffic analysis; identified and fixed critical data-exposure risks
- Executed secure code reviews in Python and Java — identified algorithmic bottlenecks and logical execution flaws in production
Network Systems Intern
@ Cisco | May 2024 – Jul 2024
- Simulated 200+ node enterprise networks with strict ACL architecture
- Analyzed raw PCAPs in Wireshark to track TCP states and isolate anomalous traffic patterns
Co-Coordinator, Xploit
@ IIIT Bhopal Cybersecurity Club | Aug 2025 – Present
- Architected and maintained scalable server infrastructure for Capture The Flag (CTF) competition environments — supporting concurrent challenge deployments under load
- Trained 700+ students in reverse engineering, vulnerability research, and offensive security fundamentals through structured workshops and live CTF competitions
03. Infrastructure & Projects
NexusDB
GitHub Repository ↗Bare-Metal Vector Engine & RAG FLAGSHIP
A custom vector database written entirely from scratch in C++17. Implements a Hierarchical Navigable Small World (HNSW) graph index for incredibly fast semantic retrieval. Features a completely air gapped RAG pipeline using local LLMs (Ollama) and manual SIMD (AVX) loop unrolling for high speed mathematical throughput.
Aegis DPI
GitHub Repository ↗Deep Packet Inspection Engine FLAGSHIP
Enterprise-grade C++17 DPI engine. 1 Reader thread → FNV-1a hash routing across 4 Load Balancer threads → 16 Fast Path workers. alignas(64) cache-line padding eliminates L1/L2 false sharing. Reverse-engineers TLS Client Hello to extract SNI without breaking encryption. Zero external dependencies — libpcap entirely bypassed via native std::ifstream binary mapping.
Concurrent Backend Architecture
A highly concurrent backend architecture built with Python and PostgreSQL. Engineered to handle strict ACID transactions for distributed, multi-tiered systems under heavy concurrency by employing robust TOCTOU elimination algorithms.
Full-Stack Secrets Manager ACTIVE BUILD
Secure full-stack secrets manager leveraging the MERN stack with JWT authentication, full 2FA implementation, and robust AES-256-CBC encryption at rest. Integrates detailed logging and enterprise-grade security within a modern, responsive Tailwind UI.
Fileless Malware Detection Engine SECURITY RESEARCH
Forensic tool detecting process-injection attacks and fileless malware by parsing raw memory dumps directly — evades disk-based AV defenses entirely. Identifies shellcode injection, hollowed processes, and anomalous memory regions.