Open to remote roles

Adversarial testing
for teams that ship.

I'm Rahil Sahu — an Application Security Engineer based in Indore, India. I run VAPT on web, API and mobile applications, validate real exploitability, and translate findings into remediation that fits engineering deadlines. I'm currently open to full-time remote roles in penetration testing and application security.

700+ validated findings
86 Critical
194 High
300+ TPRM assessments
4+ years in practice

What I test, and how I test it.

Every engagement is scoped to a recognised standard so findings translate cleanly into a remediation plan the team can act on.

Scope Standard Primary tools Deliverable
Web applications OWASP Top 10 · WSTG Burp Suite, SQLMap, Nuclei, Nessus Prioritised findings + retest
APIs (REST / GraphQL) OWASP API Security Top 10 Burp Suite, HTTPx, custom Python fuzzers Risk-scored report + secure spec review
Mobile (Android / iOS) OWASP MASVS / MSTG MobSF, Frida, Objection, JADX Static + dynamic review, hooks and PoC
Network & infrastructure PTES · NIST SP 800-115 Nmap, Nessus, Wireshark, Metasploit External + internal, hardening guidance
Cloud (Azure / AWS / GCP) CIS Benchmarks · CSA CCM SentinelOne CSPM, CloudBrute, native tooling Config review + IAM + logging validation
Security architecture Threat modelling · secure design review Data-flow & trust-boundary analysis, design walkthroughs Threats + required controls, pre-implementation
Source code review Secure SDLC · CWE Top 25 SonarQube, SonarCloud, Snyk, CodeAnt.ai Findings mapped to code, PR-ready fixes

Security architecture & design review

Engaged before code is written: reviewing proposed designs, modelling threats and defining the controls a feature needs, so architectural risk is closed at design time rather than found in a pentest.

Authentication architecture

JWT design and review across token lifecycle, expiry, refresh-token handling, signature validation, claims, key management and secure storage. MFA review covering enrollment, OTP/TOTP handling, recovery, session binding and bypass paths.

Authorization & trust boundaries

RBAC and least-privilege design, API authorization, session management. Trust-boundary and data-flow review targeting broken access control, IDOR/BOLA, privilege escalation and authentication bypass before they ship.

File handling & data protection

Upload architecture controls: file-type validation, content inspection, size limits, secure storage, access control, filename handling and prevention of malicious file execution — plus encryption and sensitive-data protection requirements.

Threat modelling

Attack-surface analysis across application components, APIs, authentication flows and data-processing workflows, with security requirements defined per feature and reusable design patterns fed back to the teams.

Beyond testing

Security programme support

Third-party risk management, ISO 27001 / GDPR / HIPAA evidence gathering, and 300+ vendor security questionnaires completed end-to-end.

SecOps & monitoring

Wazuh SIEM deployment (Docker) with VirusTotal integration, Cloudflare WAF tuning, SentinelOne EDR/CSPM operation, DefectDojo for vulnerability management.

Awareness & training

GoPhish and Evilginx phishing simulations, secure coding sessions for engineering teams, and remediation walkthroughs with developers.

Where I've done the work.

  1. May 2024 — present Indore, India

    Penetration Tester

    TeamLease RegTech Pvt. Ltd.

    • Sole information-security resource: end-to-end responsibility across programme, tooling and validation.
    • Ran VAPT across web, mobile (Android/iOS) and API surfaces; reported 86 Critical and 194 High findings against OWASP Top 10 and API Security Top 10.
    • Own security architecture and design review: threat modelling, JWT and MFA authentication design, RBAC and API authorization, file-upload controls and trust-boundary analysis — carried out at design time with product and engineering.
    • Deployed Wazuh SIEM in Docker with VirusTotal integration; operated Cloudflare WAF, SentinelOne EDR/CSPM, and DefectDojo for vulnerability lifecycle management on Microsoft Azure infrastructure.
    • Stood up secure remote access via JumpServer and OpenRport; implemented SAST/DAST with SonarQube, SonarCloud, Snyk and CodeAnt.ai.
    • Ran phishing simulations with GoPhish and Evilginx; completed 300+ TPRM vendor assessments; led secure coding training and cross-team remediation.
  2. Sep 2022 — May 2024 Indore, India

    Cyber Security Analyst

    Softude Infotech Pvt. Ltd.

    • Ran VAPT for web, mobile and API applications; primary point of contact for external engagements, retesting and remediation validation.
    • Automated secure backup, encryption and cloud transfer flows in Python across AWS services.
    • Participated in ISO 27001 audits and ensured alignment with GDPR handling requirements.
    • Ran GoPhish phishing campaigns and delivered secure-coding sessions to engineering teams.
  3. Aug 2021 — Nov 2021 Internship

    Java Developer Intern

    Collexa Technologies Pvt. Ltd.

    • Built Java applications using OOP principles and core data structures.
    • Shipped a Library Management System with modular class design and exception handling.

Things I built to make the job easier.

Side projects, all vibe-coded and all GUI-first — built so a tester (or a student) can run them without touching a terminal. The security tooling exists to exercise my own lab and authorised engagements; none of it is published as a ready-to-run weapon.

web application · education flagship

StudyFree GPT

A free education ecosystem for Indian competitive-exam aspirants, built around one rule: optimise for what a student actually learns, not for time-on-app. No AI model is hosted — students bring their own API keys and unlock only the features they want to pay for, so the platform stays free to run and free to use.

What it does

  • Built-in syllabus coverage with per-topic progress tracking.
  • AI-generated study roadmap that adapts as the student's performance data comes in.
  • Learning-pattern analysis that reshapes pace, sequencing and revision intervals per student.
  • Current-affairs feed powered by the student's own Gemini key.
  • Personal tutor mode powered by the student's own OpenAI key.
  • Two roles — student and admin — with separate surfaces and separate privileges.

How it's built

  • Bring-your-own-key model: no vendor keys on the server, no inference costs, no lock-in.
  • All data at rest encrypted; credentials hashed; student API keys stored encrypted and scoped to that student.
  • Role separation and least privilege enforced server-side, not in the UI.
  • Architected to scale from MVP to an enterprise-grade ecosystem without a rewrite.
  • bring-your-own-key AI
  • encryption at rest
  • RBAC
  • adaptive learning
  • secure-by-design
appsec tooling GUI

Bulk URL Checker

Takes a list of URLs from CSV, XLSX or TXT, hits them all in one run, and records status code, response headers and timing for each. Produces a detailed report, with optional automated PoC screenshots per URL — which turns an afternoon of manual triage into a single click.

  • csv / xlsx / txt input
  • header & status capture
  • PoC screenshots
upload testing GUI

Malicious File Generator

Embeds test payloads into arbitrary file types to exercise file-upload controls end to end — extension allow-lists, MIME and magic-byte validation, AV and sandbox scanning, and what the application does with a file once it lands. Built for authorised assessments against my own lab and in-scope targets.

  • polyglot / embedded payloads
  • content-type validation
  • OWASP file upload
upload testing GUI

RLO Filename Spoof Generator

Generates filenames carrying the Unicode right-to-left override character (U+202E) to demonstrate extension spoofing — invoice<U+202E>fdp.exe renders as invoiceexe.pdf in most file pickers and mail clients while still executing as a binary. Useful for proving that an upload filter reads the displayed name rather than the real one, and for phishing-awareness demos.

  • U+202E
  • MITRE T1036.002
  • filter bypass PoC
upload testing GUI

File Size Inflator

Pads any file type with dummy data up to a target size on demand. Written for the boring half of upload testing: finding the real size ceiling, checking whether the limit is enforced server-side or only in JavaScript, and seeing how WAF body caps, timeouts and storage quotas behave at the boundary.

  • arbitrary file types
  • exact target size
  • limit & quota testing
browser extension Chromium

Ad Blocker Extension

A request-filtering extension for Chromium-based browsers that removes roughly 80% of ad placements on the sites I use daily. Built mostly to understand the filtering pipeline from the inside — rule matching, request interception and what the browser will and won't let an extension block.

  • request filtering
  • rule matching
  • ~80% block rate
browser extension Chromium

Meeting Attendance Recorder

Captures the attendee roster of a live browser-based meeting — every participant who joined, the total head count, and how long each person actually stayed — then exports it as a clean attendance register. Replaces scrolling the participant panel and counting by hand at the end of a session.

  • join / leave timestamps
  • per-person dwell time
  • exportable register

Open source on GitHub

Everything below is public, MIT-licensed and tested — the code is the claim. Full list on github.com/RahilSahu.

ai security platform flagship

Aegis Security Platform

Scanners find; Aegis proves. It ingests Nuclei, ZAP, Burp and generic scanner output, dedupes the same issue across tools, then verifies every finding with deterministic, non-destructive checks before AI triage writes a word — the AI can never set the verification status. Evidence-grade reports in Markdown, HTML and JSON, with a retest loop that diffs fixed, persistent and regressed findings.

  • cross-tool dedupe
  • deterministic verification
  • AI triage
  • retest diffing
ai product live demo

AdaptiveChat

A chatbot that sets its own temperature for every message — facts run cold, brainstorming runs hot — behind input and output guardrails, orchestrated as a LangGraph state machine. Upload documents and answers come from your files with cited sources. Bring your own API key in the browser; nothing is stored server-side.

  • self-adjusting temperature
  • LangGraph orchestration
  • document RAG
  • bring-your-own-key
systems pure stdlib

Distributed Task Queue

A production-shaped task queue in pure Python standard library — no dependencies. Priority and FIFO leasing with owner-checked ack, fail and heartbeat, exponential-backoff retries into a dead-letter queue, delayed jobs and atomic JSON persistence with at-least-once delivery semantics.

  • priority leasing
  • dead-letter queue
  • backoff retries
  • zero dependencies
cloud security 4 clouds

Cloud Misconfiguration Scanner

Finds the exposures that turn into breaches: unauthenticated public-access checks plus offline analysers for AWS, Azure, GCP and OCI policies, security groups and firewalls. Runs against a whole fleet from one targets file and exports a JSON report.

  • AWS · Azure · GCP · OCI
  • public exposure checks
  • fleet mode
appsec tooling 10 tools

Zero-Dependency Security Tools

A shelf of small, single-purpose tools that run anywhere Python does: HTTP security headers checker, TLS certificate inspector, DNS and email-auth checker, JWT analyzer, CORS misconfiguration checker, bulk URL checker, port scanner lite and more. Every tool ships with offline unit tests.

  • zero dependencies
  • offline tests
  • MIT licensed
writing blog

AppSec Field Notes

Original notes from real engagements, written to be used: an HTTP security headers audit, a file upload testing checklist, a JWT audit walkthrough, SPF, DKIM and DMARC explained from the attacker's side, and how RLO filename spoofing hides a file's true type.

  • HTTP headers
  • file uploads
  • JWT
  • email auth

What's in the kit.

Filter by category to see the tools I use day-to-day.

  • Burp Suiteproxy · fuzzer
  • SQLMapsqli
  • Nucleitemplated scanner
  • HTTPxhttp toolkit
  • Nessusvuln scanner
  • AcunetixDAST
  • AWScloud infra
  • GCPcloud infra
  • OpenRportremote access
  • VirusTotalthreat intel
  • Nmapdiscovery
  • Wiresharkpcap analysis
  • Metasploitexploitation
  • WinPEASpost-exploit
  • AccessChkwindows acl
  • MobSFstatic + dynamic
  • Fridahooking
  • Objectionruntime
  • JADXreverse
  • Subfindersubdomains
  • Amassattack surface
  • CloudBrutecloud recon
  • HexStrike.aiai-augmented offensive
  • CodeAnt.aiai SAST
  • Venice AIuncensored llm
  • WazuhSIEM
  • CloudflareWAF
  • SentinelOneEDR · CSPM
  • DefectDojovuln mgmt
  • SonarQubeSAST
  • SnykSCA
  • JumpServerPAM
  • GoPhishphishing sim
  • Evilginxphishing sim
  • Azurecloud infra
  • Dockercontainers
  • Gitvcs
  • Pythonautomation
  • Kali Linuxworkstation

Certifications & education.

EC-Council Active

Certified Ethical Hacker (CEH)

Issued
Aug 2024
Expires
Aug 2027
Credential ID
ECC1859370462
The SecOps Group Active

Certified Cloud Security Practitioner — AWS

Issued
May 2023
Credential ID
7476871
The SecOps Group Active

Certified AppSec Practitioner (CAP)

Issued
Feb 2023
Credential ID
7040395
Purple Academy Active

Modern Log Management Fundamentals

Issued
Aug 2023
Credential ID
64d480dfcb78e8279018f9

Education

  • B.Tech, Computer Science & Engineering ADINA Institute of Science and Technology Jun 2018 — May 2022
  • Higher Secondary (PCM) Joy Campion Public School May 2016 — May 2018

Playground — try it yourself.

Four small challenges rendered entirely in your browser. No servers, no tracking, no execution of your input — each check is client-side and read-only. Solve them for practice, or peek at the hint if you get stuck.

#01 HIGH

Bypass the sanitiser

A junior dev wrote this "XSS filter". It strips <script>, onerror=, onload=, and javascript:. Submit a payload that would still execute JavaScript.

function sanitise(input) {
  return input
    .replace(/<script>/gi, '')
    .replace(/onerror=/gi, '')
    .replace(/onload=/gi,  '')
    .replace(/javascript:/gi, '');
}

#02 EASY

Find the flag

A flag in the format FLAG{...} is hidden somewhere in this challenge card. Not in visible text. Inspect the DOM (browser DevTools → Elements) and submit it.

There is nothing to see here — keep looking with the right tool.

#03 MED

Read the JWT

A production app issues this token. Decode the payload and submit the role claim value (lowercase).

eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzdWIiOiJyYWhpbCIsInJvbGUiOiJndWVzdCIsImV4cCI6MTc5MzQ1Njc4OX0.k9F3fD7pQr4wJ8mV2nA6xL1zK5tHbG0uYcRvE8sMdIQ

#04 MED

Peel the encoding

This ciphertext was base64-encoded twice, then ROT13'd on the outside. Recover the flag.

Hzg4DyVmqQOwoJfjJwWIqR0lAJcvZyWjLz1wqTWRHwIAZ0c6MyR9CD==

Every check runs locally. Your input never leaves this tab — verify in DevTools → Network.

Learn offensive security — free.

Curated free platforms that carried me through 4+ years of practice. Everything below is either free-tier or fully free. Pick one, don't jump between five.

web · foundational free

PortSwigger Web Security Academy

The definitive free web-hacking curriculum. Interactive labs mapped to OWASP Top 10. Start here.

portswigger.net/web-security →
guided rooms freemium

TryHackMe

Beginner-friendly guided rooms with structured learning paths. Great for building foundational skills.

tryhackme.com →
CTF · boxes freemium

HackTheBox

Real vulnerable machines. Steeper curve than THM — retire boxes are free, active boxes require VIP.

hackthebox.com →
web · code review freemium

PentesterLab

Deep-dive exercises on specific bug classes: SSRF, JWT, deserialisation, race conditions. Excellent code review track.

pentesterlab.com/exercises →
standards free

OWASP Top 10 & WSTG

Read the standard you'll be assessed against. Testing Guide (WSTG) is a full methodology playbook.

owasp.org →
bug bounty prep free

Hacker101 CTF

HackerOne's free CTF. Solve enough to earn private-program invites on the platform. Video lessons alongside.

ctf.hacker101.com →
linux · shell free

OverTheWire Wargames

SSH into games. Bandit teaches Linux fundamentals; Natas covers server-side web. Nothing beats it for shell fluency.

overthewire.org/wargames →
offline labs free

VulnHub

Downloadable vulnerable VMs. Run them locally — no network dependency, no VPN, unlimited resets.

vulnhub.com →
blue team · DFIR freemium

CyberDefenders

Blue-team side of the house: forensics, incident response, threat hunting. Complements offensive practice.

cyberdefenders.org →

AI × Security.

Where I'm spending research time: attacking LLM-integrated applications, and using AI to accelerate offensive workflows. Both sides matter — the attack surface is new, and so is the tooling.

Working note: Client-side prompt-injection demos are deliberately kept off this site — a public payload gallery can be scraped and used against systems it wasn't intended for. Reach out for a scoped walkthrough.

Hire me