M.S. Electrical & Computer Engineering โ€” Northeastern University

Sreenidhi Ramani

Cybersecurity Graduate Student | SOC & Blue Team | GRC & Pentesting | Threat Intelligence & Security Engineering

about.txt

whoami

I study how embedded systems fail before an attacker finds out first. Most of my time right now goes into fuzzing firmware on a seizure-detection wristband โ€” but the throughline across everything I do is the same question: how does this actually break?

I'm a Cybersecurity Graduate Student at Northeastern University conducting firmware security research and building automated vulnerability discovery tooling. Previously, I worked across offensive and defensive security in Mumbai, running penetration tests and supporting incident response for live security events.

Today, I focus on SOC and blue team work, penetration testing, and embedded security, alongside CTF competition with Northeastern's NUSec team. Full experience details below.

Currently seeking cybersecurity Full-time opportunities for 2026

Sreenidhi Ramani

// experience

Research Assistant, CactiLab
Northeastern University
May 2026 โ€“ Present
  • Run an automated vulnerability discovery pipeline against NeurAlert (ARM Cortex-M, FreeRTOS) using LibFuzzer and Boofuzz in a custom QEMU-based emulated environment โ€” 23+ findings so far
  • Built Python tooling that generates SARIF reports directly from fuzzer crash logs, and maintain the structured firmware dataset behind the lab's ongoing research
Data and Event Assistant, Academic Ceremonies
Northeastern University
Jan 2026 โ€“ Present
  • Manage data integrity for Commencement operations serving 35,000+ graduates across 8 colleges, handling sensitive student records with strict attention to confidentiality
  • Coordinate inventory and cross-team logistics for large-scale, high-volume events under tight timelines
Cybersecurity Analyst Intern
Cyber Secured India โ€” Mumbai
Jul 2024 โ€“ Dec 2024
  • Ran vulnerability scans and pentests across 8โ€“12 web/mobile apps with Nmap, Metasploit, Burp Suite, and OWASP ZAP, identifying 20+ vulnerabilities including SQLi, XSS, and insecure auth
  • Monitored network traffic and logs with Splunk, Wireshark, and Snort, investigating 100+ security events; ran incident response on 10+ cases using Nikto and Autopsy
Cybersecurity Intern
VIEH Group Private Limited
Feb 2024 โ€“ Apr 2024
  • Performed web application pentesting with Burp Suite, Nmap, and Metasploit, uncovering 6+ critical vulnerabilities including insecure APIs and weak encryption
  • Wrote Python scripts for network traffic analysis and automated anomaly detection, improving detection accuracy by 30%
Professional Development
Self-Directed
Ongoing
  • Working through SOC and Blue Team labs on TryHackMe and solving CTF challenges on PicoCTF to sharpen detection and incident response skills
  • Preparing for the CompTIA Security+ certification

// skills

Firmware & Embedded
LibFuzzerBoofuzzQEMUARM Cortex-MFreeRTOSASANSARIFESP8266ArduinoMQTT
Offensive Security
Burp SuiteMetasploitNmapOWASP ZAPNiktoHydraSQLmapGobusterKali Linux
Defensive / SOC
SplunkWiresharkSnortAutopsyMITRE ATT&CKSIEMAlert TuningIncident Response
Digital Forensics
AutopsyVolatilityBinwalkJohn the Ripper
GRC
NIST SP 800-53NIST CSF 2.0SOC 2ISO 27001Risk RegistersFIPS 199
Languages
PythonCC++BashSQL

// education

M.S. Electrical & Computer Engineering โ€” Expected Dec 2027
Northeastern University, Boston, MA
Concentration: Computer Systems & Software
B.E. Electronic & Telecommunication Engineering, Honors in Cybersecurity
Dr. D.Y. Patil Institute of Engineering, Management and Research, Pune

// extracurricular

CTF Member
NUSec, Northeastern University
  • Compete in team-based security competitions on TryHackMe and PicoCTF
Technical Head
IEEE Student Branch
2023 โ€“ 2024
  • Directed logistics for workshops and hackathons serving 100+ students; managed registrations, scheduling, resource allocation, and communications to ensure smooth operations
Joint Secretary
Talk Club
2023 โ€“ 2024
  • Handled correspondence, attendance records, and scheduling of weekly seminars; managed social media outreach and grew participation by 30% through consistent engagement
Junior Public Manager
Entrepreneurship Development Cell (EDC)
2021 โ€“ 2022
  • Coordinated E-Summit'22, a 3-day entrepreneurship event; oversaw logistics, participant management, and communications for 500+ attendees
Event Team Head
College Cultural Fest
2021 โ€“ 2022
  • Led end-to-end coordination for multi-day events with 500+ attendees; managed vendor communication, logistics, and on-ground operations across the full event lifecycle

Also: 8 years of Bharatanatyam training, and basketball.

ls -la ./projects/

Projects

Firmware, offensive security, and GRC tooling โ€” filter by track.

Systems Security

ShellGuard โ€” POSIX Shell with Exploit Research & NIST Controls

Dec 2025 โ€“ Feb 2026

A POSIX shell built in C (fork, execvp, pipe, dup2, waitpid, multi-stage pipelines) with an audit logging module mapped to NIST SP 800-53 AU-2/AU-3/AU-9/AU-12. Includes working exploit demos for command injection (CWE-78) and UID spoofing via NFSv4 Kerberos AUTH_SYS downgrade (CWE-287/290), each paired with a hardened fix.

CLinuxNIST 800-53
Offensive Security

SOC Home Lab โ€” Active Directory Attack & Detection

Dec 2025 โ€“ Feb 2026

A self-built Active Directory environment for practicing common AD attack chains and the corresponding blue-team detections, from initial access through privilege escalation.

Active DirectoryDetection Engineering
Post-Quantum / IoT

Adaptive Reconciliation-Based PQC Key Exchange for IoT

with Cherisma Tanduru ยท EECE 5155

A Module-LWE based key exchange protocol for secure IoT communications, built with an efficient error-reconciliation scheme, performance benchmarking, and NS-3 network simulation.

C++Module-LWENS-3
Fraud Detection

AI-Powered Elder Fraud Detection Pipeline

A detection pipeline for elder financial exploitation that flags account takeover, romance and grandparent scams, tech support scams, and phishing-enabled credential resets before funds leave the account โ€” applying the same class of unsupervised anomaly detection used in enterprise SOCs to a population that rarely has access to that tooling.

Pythonscikit-learnIsolation Forestpandas
GRC

SecureForGood โ€” AI-Powered Compliance Auditing for Nonprofits

Automates enterprise-grade security auditing for organizations that can't afford it โ€” scans Windows security configurations, maps findings to NIST SP 800-53 controls using GPT-4, identifies compliance gaps with risk scores, and generates plain-English PDF and JSON reports for both technical teams and nonprofit leadership. Built with a LangChain agentic orchestration layer across six phases.

PythonLangChainGPT-4SQLite FTS5
Networking

Portable DDoS Attack Simulation Tool

A portable DDoS attack simulator built on ESP8266 โ€” an educational cybersecurity research project for network security training and controlled penetration testing.

C++ESP8266
Networking

Packet Sniffer Tool

A low-level network packet sniffer using raw sockets in Python, parsing Ethernet, IPv4, TCP, UDP, and ICMP directly off the wire.

PythonRaw Sockets
Published ยท IJRASET 2022

Cloud Storage Connected Security Safe

A physical security safe concept connected to cloud storage for remote monitoring and access control, published in the International Journal for Research in Applied Science & Engineering Technology.

Read the paper โ†’
Published ยท IJERT / ICSCSET 2025

PrepPal โ€” Peer Matchmaking to Combat Study Isolation

A peer matchmaking platform tackling study isolation among students, published in IJERT Vol. 14 (April 2025) and presented at ICSCSET 2025.

Read the paper โ†’

cat blog/*.log

Writeups

Notes from the lab and from CTFs โ€” drafts in progress, published as I finish them.

Published

Welcome to My Blog click to expand โ–พclick to collapse โ–ด

Welcome to my corner of the internet. This is where I'll be writing up what I'm actually working on โ€” SOC and blue team practice, pentesting, GRC, and whatever I pick up along the way in CTFs.

What to Expect

  • SOC & Blue Team Writeups โ€” Walkthroughs from TryHackMe rooms and PicoCTF challenges as I work through them
  • Pentesting & Offensive Security โ€” Notes from web app and network engagements, tooling, and methodology
  • GRC & Risk โ€” Working through frameworks like NIST CSF and SOC 2, and what actually applying them looks like
  • Career & Learning โ€” What's actually helping (or not) in my M.S. ECE journey and CompTIA Security+ prep

Why a Blog?

Writing about a topic forces me to actually understand it โ€” it's easy to follow a checklist and call it done, harder to explain the "why" behind it clearly enough for someone else to follow. I'd rather build that habit in public than not.

Sample: Quick Recon & Triage Cheat Sheet

A few commands I reach for constantly during the recon and initial-triage phase of an engagement or lab:

# Quick host discovery across a subnet
nmap -sn 192.168.1.0/24

# Full port scan with service and script detection
nmap -sV -sC -p- -oN full_scan.txt target.com

# Pull recent failed logon events for triage (Windows)
Get-WinEvent -FilterHashtable @{LogName='Security'; ID=4625} -MaxEvents 50

# Quick tail of auth logs on a Linux host
tail -f /var/log/auth.log | grep -i "failed\|invalid"

Tip: always get proper authorization before scanning or testing any network or system that isn't yours.

Stay Tuned

More writeups coming as I finish them. In the meantime, feel free to check out my projects or get in touch.

Published

Writing My First Sigma Rule click to expand โ–พclick to collapse โ–ด

Notes from working through TryHackMe's SOC Level 2 path โ€” on the gap between knowing what a log looks like and knowing what to write a detection rule for.

Part of my SOC Level 2 roadmap right now is Splunk BOTS and Sigma rule writing. I'd read plenty of Sigma rules before, but reading one and writing one from scratch turned out to be two very different skills.

The Setup

Sigma is a generic, SIEM-agnostic format for writing detection logic โ€” you describe the pattern you're looking for once, and it compiles down to whatever your backend actually speaks (Splunk SPL, Elastic, Sentinel, etc.). The pitch is "write it once, detect everywhere." The reality, at least for a first attempt, is "figure out exactly which log fields actually fire, then write it once."

What Tripped Me Up

  • Field names aren't universal โ€” the field you'd expect (like a process name) doesn't always match across log sources the way the Sigma spec examples suggest
  • Too broad vs. too narrow โ€” my first draft rule matched almost nothing; my second matched way too much. Tuning a detection is its own skill, separate from writing the YAML
  • Context matters more than the match โ€” a rule firing correctly is only half the job; knowing whether that activity is actually suspicious in context is the other half

Sample: A Basic Sigma Rule Skeleton

Here's roughly the shape of a simple detection rule, structured to flag suspicious PowerShell execution:

title: Suspicious PowerShell Encoded Command
status: experimental
logsource:
  category: process_creation
  product: windows
detection:
  selection:
    Image|endswith: '\powershell.exe'
    CommandLine|contains:
      - '-enc '
      - '-EncodedCommand'
  condition: selection
level: medium

Tip: start every rule with a narrow, high-confidence selection, then widen it only after you've confirmed what normal traffic looks like in your own environment.

Next Up

Still working through the rest of the SOC Level 2 path alongside Splunk BOTS data sets โ€” more detection engineering notes as I go.

Published

Breaking Into AI Security: Notes from TryHackMe's AI Security Path click to expand โ–พclick to collapse โ–ด

TryHackMe rolled out a dedicated AI Security path recently, and it's become a genuinely good on-ramp into a field that didn't really have structured hands-on material a year ago.

AI security is a weird field to break into right now โ€” there's a lot of theory floating around (prompt injection is everywhere in the news), but not a lot of structured, hands-on material. TryHackMe's new AI Security path is one of the first places I've found that actually walks through it room by room instead of just talking about it.

The Path So Far

  • The Building Blocks of AI โ€” a short, interactive room covering the fundamentals: how models are trained, what "inference" actually means, and the basic vocabulary the rest of the path assumes you know
  • LLM Security โ€” frames LLMs as an attack surface in two halves: data-based threats (membership inference, training data extraction) and model-based threats (the model's own behavior being manipulated or abused)
  • Securing AI Systems โ€” the most practical room so far. It has you map out an AI assistant's architecture, identify where the OWASP LLM Top 10 and MITRE ATLAS attack categories actually apply, and design trust boundaries around it
  • AI Threat Modelling โ€” covers prompt injection, data poisoning, and excessive agency โ€” the last one being the "confused deputy" problem, where an AI system is granted more permissions or autonomy than it should safely have

The Concept That Clicked: Excessive Agency

Most AI security writing focuses on prompt injection because it's flashy, but excessive agency is the one that actually worries me more. It's the classic confused-deputy problem wearing a new costume: an AI agent with legitimate access to internal APIs or documents can be tricked into misusing that access on someone else's behalf โ€” not because the model was "hacked," but because it was never scoped tightly enough in the first place. The fix isn't a smarter model, it's smaller permissions.

Reference Material Worth Bookmarking

  • MITRE ATLAS (atlas.mitre.org) โ€” the ATT&CK-style framework for adversarial threats against AI systems; good to skim just to recognize the terminology
  • OWASP GenAI Security Project (genai.owasp.org) โ€” home of the OWASP LLM Top 10, free to read, no account needed

Tip: if you're coming from traditional appsec, map each new AI-specific threat back to something you already know โ€” prompt injection rhymes with injection attacks, excessive agency rhymes with least-privilege violations. The underlying security principles haven't changed; the attack surface has.

Next Up

Working through the rest of the path, then pairing it with the OWASP LLM Top 10 directly against a project to see how the threat categories actually show up in practice.

./contact

Let's talk

Looking for security internships and co-ops across SOC, GRC, security engineering, and hardware/firmware security. Reach out โ€” I'd love to hear what you're working on.

Email ramani.sr@northeastern.edu
GitHub /Sreenidhi15
TryHackMe Ciph3r15
Location Boston, MA