~/ Breaking things to build them better.

Cybersecurity engineer · NYU Tandon MS ’26 · Brooklyn, NY

Research Coordinator at NYU’s OSIRIS Lab. I build and break low-level systems: a post-quantum secure channel in C, a race-condition fuzzer for the Linux network stack, a hardened xv6 kernel, and a CSAW CTF 2025 challenge.

View work Résumé (PDF) lp2989@nyu.edu

Open to security roles

Now

Research Coordinator, OSIRIS Lab leading RaceFuzz Lind project researcher CSAW CTF 2025 challenge author

~/work

Selected work

All 11 projects
channel.design
  1. kem ML-KEM-768 or HQC-192
  2. sig ML-DSA-65 or SLH-DSA (SPHINCS+)
  3. kdf HKDF-SHA256
  4. aead AES-256-GCM
  5. test one Docker harness · NIST PQ Category 3

Post-quantum cryptography · 2026 · NYU CS6903

Kyber Crystals: one secure channel, two post-quantum families

A research-grade post-quantum secure channel in C, implemented twice behind one pluggable backend: ML-KEM-768 (Kyber) with ML-DSA-65 (Dilithium), and lattice-free HQC-192 with SLH-DSA (SPHINCS+).

Both backends sealed with AES-256-GCM and HKDF-SHA256, then attacked and benchmarked in a unified Docker test harness at NIST PQ Category 3.

  • C
  • liboqs
  • ML-KEM
  • SPHINCS+
  • Docker

Case study Source on GitHub Write-up

racefuzz.pipeline
  1. gen corpus generation in Go
  2. mut coverage-guided semantic mutation
  3. run concurrent paths into the Linux network stack
  4. find race-condition detection heuristics

Kernel fuzzing · 2026 · OSIRIS Lab

RaceFuzz: hunting race conditions in the kernel network stack

A semantic fuzzing framework that targets race conditions in the Linux kernel’s network stack, combining coverage-guided mutation, custom Go-based corpus generation and concurrency analysis.

The OSIRIS Lab research project I lead, one of the nine projects I coordinate.

  • Go
  • C
  • Kernel fuzzing
  • TCP/IP stack

Case study Write-up

robustness.experiment
  1. data MNIST
  2. train PGD adversarial training (Madry et al., 2018)
  3. attack PGD · FGSM · C&W via CleverHans and ART
  4. port TensorFlow 1.x → 2.x

Adversarial ML · 2025 · NYU Tandon

Adversarial robustness in deep neural networks

Implemented PGD adversarial training on MNIST to validate Madry et al. (2018), re-engineering the TensorFlow 1.x code for 2.x so CleverHans and ART experiments could run.

About 90% accuracy against PGD and FGSM attacks, with weaknesses identified under C&W attacks and distribution shift.

  • Python
  • TensorFlow
  • CleverHans
  • ART

Case study Source on GitHub

~/research

Research coordination at OSIRIS Lab

OSIRIS Lab

Since February 2026 I’ve been Research Coordinator at OSIRIS Lab, NYU’s cybersecurity research group, guiding nine active projects across kernel exploitation, AI security, embedded systems and CTF infrastructure. Each project mentors students across skill levels.

  1. Semantic fuzzing for kernel network-stack race conditions Lead: Love Kush Go · C Advanced

    Developing a semantic-aware fuzzer to detect race conditions in the Linux kernel’s network stack using coverage-guided mutation and concurrency analysis.

  2. MCP server for binary exploitation Lead: Jesse Python · Pwn Beginner

    Building an MCP (Model Context Protocol) server that assists in binary exploitation workflows, automating gadget finding, ROP chain construction, and exploit generation.

  3. LLM-based CTF solver with reasoning extension Lead: Prajesh Python · LLM Intermediate

    Augmenting large language models with chain-of-thought reasoning to autonomously solve CTF challenges across crypto, pwn, and web categories.

  4. PUF-based authentication protocol Lead: Vaughn Hardware · C Advanced

    Designing a physically unclonable function (PUF) based authentication protocol for resource-constrained embedded devices resistant to side-channel and replay attacks.

  5. Secure ADS-B framework Lead: Aaditya SDR · Python Intermediate–advanced

    Developing a secure Automatic Dependent Surveillance–Broadcast (ADS-B) framework to authenticate aircraft transponder signals and defend against spoofing and injection attacks.

  6. ROS 2 and RTOS security Lead: Prinn ROS 2 · C++ Intermediate

    Analyzing security vulnerabilities in ROS 2 robotic middleware and real-time operating systems (RTOS), with focus on privilege escalation and message injection attacks.

  7. AI privacy exploitation in IoT devices Lead: Nidhi Python · IoT Intermediate

    Investigating how AI models embedded in IoT devices can be exploited for privacy leakage through model inversion, membership inference, and side-channel analysis.

  8. Pastebin data leak analysis Lead: Maya Python · OSINT All levels

    Large-scale scraping and analysis of Pastebin and similar platforms to identify leaked credentials, API keys, PII, and sensitive organisational data for responsible disclosure.

  9. Evolving CTFs for AI Lead: Kim Python · AI Intermediate

    Designing adaptive CTF challenges that evolve in difficulty and category based on an AI solver’s capabilities — creating a co-evolutionary arms race between AI and human-designed puzzles.

~/writing

Writing

All writing

  1. Jul 2026 · Post-quantum

    Building the same secure channel twice

    Why I implemented one encrypted channel with two different post-quantum families, and what attacking my own protocol taught me.

  2. Jul 2026 · Kernel

    Hunting race conditions in the kernel network stack

    Why race conditions are the bugs fuzzers are worst at finding, and what semantic fuzzing means. Notes from RaceFuzz.

  3. Jul 2026 · CTF

    What writing a CSAW CTF challenge taught me

    Solving CTF challenges makes you a better attacker. Writing one makes you a better engineer.

~/about

About

Love Kush Pranu
GPA 3.9/4.0 · NYU Tandon MS ’26

I’m a cybersecurity engineer pursuing my MS at NYU Tandon School of Engineering (GPA 3.9/4.0), expected May 2026. I graduated from NIT Calicut with a B.Tech in EEE, where circuits and systems sparked a deep passion for digital security.

As Research Coordinator at OSIRIS Lab I manage nine active research projects spanning kernel fuzzing, adversarial ML, CTF tooling and IoT security. I’m also a CTF Team Member and help run competitions like CSAW; I authored one of the CSAW CTF 2025 challenges. I like going deep, from building a post-quantum secure channel in C to modifying the xv6 kernel with ASLR and W^X protections.

Off the keyboard I run an interview series and keep a page of hobbies.

Research projects coordinated
9
Certifications & credentials
23+
GitHub repositories
61

Highlights

  • MITRE eCTF 2025: reached the Attack Phase of one of the most prestigious embedded security competitions in the US.
  • Authored a challenge for CSAW CTF 2025, designed, built and deployed for thousands of players.
  • Co-ran the CSAW CTF 2025 Qualifiers with OSIRIS Lab: challenge hosting, platform infrastructure and moderation across dozens of regions.
  • Organised Hack@NYU at Microsoft NYC: logistics, technical mentorship and stakeholder coordination.
  • Published research on adaptive distance relay coordination at NIT Calicut, improving power system reliability by 15%.

Certifications

  • CompTIA Security+
  • CySA+
  • IBM Cybersecurity
  • MITRE eCTF 2025
  • CCNA (in progress)
  • eJPT (in progress)

Toolbox

Security
Penetration testing, vulnerability assessment, incident response, IDS/IPS, firewalls (Cisco ASA), VPN/IPsec/VLAN
Systems
Python, C, C++, Rust, Go, Bash, Linux (Ubuntu, Kali), xv6, Docker, SDL2
Networks
TCP/IP, BGP, OSPF, DNS/DNSSEC, Wireshark, Nmap, Cisco IOS, iptables, Ansible
Cloud & ML
AWS (EC2, S3, IAM), TensorFlow, PyTorch, CleverHans, ART, CI/CD security, zero trust

GitHub LinkedIn CTFtime team WeChall

~/experience

Experience

Résumé (PDF)

  1. 2026 – now

    Research Coordinator, OSIRIS Lab

    New York University · Feb 2026 – present · New York City (hybrid)

    Managing 9 active security research projects across kernel fuzzing, adversarial ML, IoT security, and CTF tooling. Also a CTF Team Member since Jan 2025 — competed globally, co-ran CSAW CTF 2025 and organised Hack@NYU at Microsoft NYC.

  2. 2025 – now

    Lind Project Researcher

    New York University · Jan 2025 – present · New York City

    OS-level security research using Rust, contributing to the Lind sandbox project — a restricted POSIX environment to safely execute untrusted code via WebAssembly.

  3. 2023 – 2025

    Cyber Security Researcher, Pentadots Solutions

    Full-time · Remote · Oct 2023 – Sep 2025

    Executed red-team adversarial simulations across enterprise environments improving defense efficiency by 30%. Architected zero-trust network security frameworks with microsegmentation and least-privilege access controls.

  4. 2022

    E-Drive System Security Intern, Robert Bosch India

    Coimbatore, India · Jun – Jul 2022

    Penetration testing and vulnerability assessment on embedded E-Drive automotive systems. Security analysis of CAN bus communications and firmware hardening.

  5. 2021

    Computer Vision Research Intern, Continental AG

    Bangalore, India · Jan – Jun 2021

    Developed and evaluated computer vision models for automotive safety. Researched adversarial robustness of real-time detection pipelines in safety-critical environments.

  6. 2019 – 2023

    B.Tech, Electrical & Electronics Engineering, NIT Calicut

    National Institute of Technology, Calicut · CGPA 8.2/10

    Published research on adaptive distance relay coordination (15% reliability improvement). ISTE Joint Secretary, IEEE IAS/CAS member, Placement Cell Representative, Powerlifting Team.

~/contact

Let’s secure something great together.

Whether you have a security challenge, research collaboration, CTF opportunity, or just want to talk, I’m always open to interesting conversations. The fastest way to reach me is email; I usually reply within 24 hours.

GitHub
github.com/lutherleo
LinkedIn
linkedin.com/in/lovekush-pranu
Medium
medium.com/@lovekush.pranu
Phone
929‑414‑7472
Location
Brooklyn, New York · open to remote