Security Engineer & Researcher

Zion Boggan

I work across security operations, vulnerability research, and offensive testing — building and proving the tools that detect, investigate, and remediate real-world threats. SOC analyst by day, security researcher the rest of the time.

CybersecuritySecurity OperationsSecurity AnalystVulnerability ResearchPenetration / Security TestingTechnical Security & SupportAI Security
Memphis, TN · U.S. Citizen · Open to relocation· 00:00:00 CT
Scroll to explore
01

By the numbers

Measurable experience and credentials, at a glance.

2+ yrs
Hands-on security operations
150–300
Alerts triaged per shift
100+
Remediations validated
~35%
Fewer false-positive escalations
20+
Projects in reproducible research
6
Industry certifications
Certifications
CompTIA Security+ (SY0-701)Microsoft SC-200 (Security Operations Analyst)Microsoft AZ-104 (Azure Administrator)Microsoft AZ-900 (Azure Fundamentals)SentinelOne Incident ResponderCompTIA CySA+ (scheduled Jun 2026)
03

Professional Experience

Where I've done the work.

Jul 2024 – Present

SOC Analyst I · Cyber Guards

Memphis, TN

Front-line security operations at a managed security provider, supporting federal control baselines.

  • Monitor and triage 150–300 alerts per shift across Splunk, Microsoft Sentinel, SentinelOne, and Stellar Cyber.
  • Investigated ransomware intrusions end-to-end, producing timeline analysis and evidence that supported a successful cyber-insurance claim.
  • Validate vulnerabilities on Windows and Linux, map them to MITRE ATT&CK and NIST controls, and write remediation guidance.
  • Tracked 100+ remediation actions at a 90%+ on-time patching rate, and cut false-positive escalations ~35% through detection tuning.
Apr 2026 – Present

Independent Security Researcher · Self-employed

Memphis, TN

Coordinated vulnerability disclosure and reproducible security research.

  • Run coordinated disclosure on Bugcrowd and HackerOne across managed bug-bounty programs.
  • Work from source-code analysis, protocol review, and reproducible proof-of-concept development.
  • Findings span databases, cryptographic libraries, blockchain consensus, and authorization-bypass chains.
  • Hold every result to a strict reproduce-before-report standard.
Jun 2023 – Jun 2024

Relationship Banker · Bank of America

Memphis, TN

Customer-facing role with daily security, compliance, and identity-verification duties.

  • Performed daily security and compliance checks and authenticated customer identities with multi-factor procedures at full policy compliance.
04

Technical Skills

Grouped by what they're used for.

SIEM & Detection Engineering

Splunk, Microsoft Sentinel, SentinelOne, Stellar Cyber, Wazuh, Elastic, KQL, Sigma, detection-as-code, alert triage, MITRE ATT&CK mapping

Incident Response & Forensics

Ransomware investigation, forensic timeline analysis, memory forensics (Volatility 3), IOC documentation, evidence packaging

Vulnerability Research & Testing

Source-code review, proof-of-concept development, coordinated disclosure (Bugcrowd, HackerOne), Qualys VMDR, Nessus, OpenVAS, remediation tracking

Purple Team & Adversary Emulation

Atomic Red Team, MITRE Caldera, detection validation, coverage analysis, file-integrity monitoring

Secure SDLC / DevSecOps

CI security pipelines, Semgrep (SAST), gitleaks, pip-audit, Sigstore / Cosign signing, SBOM (syft), grype, Kyverno policy

Endpoint & Identity

CrowdStrike Falcon, Microsoft Defender for Endpoint, Microsoft 365 Defender, Tanium, Okta, Mimecast

Windows & Linux

Windows 10/11 & Server, Active Directory, Group Policy, PowerShell, Sysmon; Debian, Ubuntu, RHEL, systemd, Bash, OS hardening

Cloud & Networking

Microsoft Azure (Sentinel, Defender for Cloud, IAM), TCP/IP, DNS, routing, firewalls, VLANs, Wireshark, WireGuard, pfSense, Palo Alto, Fortinet

Compliance & Risk

NIST SP 800-171 & 800-53, CMMC, DFARS 252.204-7012, STIG hardening, control-gap assessment, policy & SOP authoring

Scripting & Languages

Python, PowerShell, Bash, KQL, Rust, FastAPI, REST API integration, Git

05

The Evidence Ledger

The full index — every project I've designed, built, and run. Click any row for the architecture, code, and evidence.

No. / Project · 21 projects Class · Year · Status
01
Flywheel
An autonomous vulnerability-research platform: a 7-agent decision graph that maps a target, hypothesizes bugs, drafts exploits, and only reports findings a deterministic validator could reproduce.
VULN RESEARCH
2025
ACTIVE
02
CYCLOPS
An autonomous vulnerability-research engine: reproduced findings across 20+ open-source targets using a strict reproduce-before-report discipline. ASan crashes, live SQL rows, and executed command chains are the only valid proof.
VULN RESEARCH
2026
RESEARCH
03
GEMINI Malware-Research Lab
A self-contained malware-analysis lab: static-triage pipeline, ransomware detonation in a sealed VM, live memory forensics, and AES key recovery from a 4.13 GB ransomware memory dump.
VULN RESEARCH
2026
SHIPPED
04
JWT Differential Fuzzer
Feeds identical tokens into 5 JWT libraries at once and flags any disagreement. A verifier split is an auth-bypass primitive.
VULN RESEARCH
2026
GHSA PENDING
05
Security Research Notebook
Public writeups and methodology notes from independent vulnerability research on HackerOne and Bugcrowd, organized by target class, with reproducible PoCs and disclosure timelines.
VULN RESEARCH
2026
ONGOING
06
CTI Detection Automation
Pulls live threat-intel from five feeds, deduplicates across sources, maps to ATT&CK, generates Wazuh detection rules, and holds every rule behind a human approval gate before anything reaches the SIEM.
DETECTION ENG
2025
SHIPPED
07
Detection-as-Code
Ten ATT&CK-mapped Sigma rules that compile cleanly to Splunk SPL, Elastic ES|QL, and Microsoft Sentinel KQL. Linted and tested in CI, behaviourally validated by Atomic Red Team before promotion.
DETECTION ENG
2025
SHIPPED
08
SOC Automation Lab
Wazuh detection, Shuffle SOAR, and TheHive case management wired into one pipeline. An alert fires, the indicator gets enriched, a case opens, and the analyst channel gets pinged, all before anyone clicks anything.
DETECTION ENG
2025
SHIPPED
09
Purple-Team Lab
Atomic Red Team adversary emulation against an instrumented Ubuntu endpoint: six ATT&CK techniques executed, six detections confirmed in Wazuh. The validation half of the detection-as-code pipeline.
DETECTION ENG
2025
SHIPPED
10
Secure CI/CD Pipeline
A GitHub Actions pipeline that gates every push on four parallel security checks (Semgrep SAST, gitleaks secret scan, pip-audit dependency audit, and ruff with security rules) before code is allowed to merge.
DETECTION ENG
2025
SHIPPED
11
CI/CD Supply Chain Security
Keyless Cosign signing with GitHub OIDC identity, SPDX SBOM attestation, and a Kyverno admission policy that refuses any unsigned or unattested image at the cluster boundary.
DETECTION ENG
2025
SHIPPED
12
Perseus
A 20-agent AI orchestration platform that routes natural language commands to specialized workers across a self-hosted Proxmox homelab, with per-request cost tracking and a three-tier SSH safety model.
AI / AUTOMATION
2026
ACTIVE
13
claude-dispatch
HMAC-signed, filesystem-mediated job dispatch so two agent sessions on different hosts can hand work to each other. No broker, no open ports, no shared terminals.
AI / AUTOMATION
2026
ACTIVE
14
gpu-cpu-mutex
Two shell scripts (a GPU mutex and a CPU/RAM counting semaphore) that let multiple independent processes share one GPU and a bounded CPU budget using nothing but flock.
AI / AUTOMATION
2026
SHIPPED
15
Hermes
Autonomous Kalshi weather-market trading bot: 31-member NWS ensemble forecasts into Gaussian bracket probabilities, Kelly sizing, a Claude Sonnet veto gate, and a Discord command interface.
QUANT
2026
ACTIVE
16
Prediction Market Bot Post-Mortem
Honest post-mortem of a Kalshi weather-market bot that lost $160 in 138 trades. Includes the actual dataset, a walk-forward eval harness, and the three cascading misdiagnoses that preceded the right answer.
QUANT
2026
POSTMORTEM
17
Pitch Tracker CV
Real-time CV aim-assist for offline MLB The Show 26: a YOLO-trained ball detector plus classical PCI tracker that predicts where each pitch will cross the plate and nudges the controller via a Titan Two adapter.
COMPUTER VISION
2025
SHIPPED
18
DeckBound
A social idle formation auto-battler for Roblox: 33 original pixel-art champions, a server-wide world boss, shared co-op runs, and a headless build pipeline that shipped to v181.
GAME DEV
2026
LIVE
19
Seven Star Collective
Full e-commerce and coaching platform for a fitness brand: merch store, session booking with calendar, workout challenge system, and Stripe checkout. Built in React/TypeScript on Supabase.
PRODUCT
2026
LIVE
20
Matcha Bloom
A private iOS matcha journal and Portland cafe map with quick-add logging, shareable Matcha Cards, monthly Bloom recaps, and 25 pre-seeded PDX cafe pins. Built in Flutter.
PRODUCT
2026
SHIPPED
21
WireGuard iOS Kill Switch Generator
A single Python script that wraps any WireGuard .conf file into an iOS .mobileconfig profile with IncludeAllNetworks and OnDemand auto-connect. These are the two kill-switch settings the WireGuard iOS app's UI does not expose.
PRODUCT
2026
SHIPPED
06

Lab Notes

Security research writeups — root-cause analysis with reproducible proof. Nothing is called a finding until it reproduces.

01
AXIS OS dnsupdate.cgi Delete Path Missing Input Validation
The add path validates all input through a dedicated function that rejects shell metacharacters. The delete path skips it entirely, passing raw user input into an nsupdate heredoc.
WRITEUP
2026
PUBLISHED
02
AXIS OS httptest.cgi SSRF via IPv6-Mapped Loopback Bypass
httptest.cgi blocks 127.0.0.1 and ::1 but treats [::ffff:127.0.0.x] as a routable address, bypassing loopback validation and reaching privileged internal Apache VirtualHosts.
WRITEUP
2026
PUBLISHED
03
AXIS OS ONVIF RTSP-over-WebSocket Endpoint Missing Authentication
Two WebSocket-to-RTSP proxy endpoints, functionally identical, sit side by side in Apache config. One has a Require directive. The ONVIF variant does not.
WRITEUP
2026
PUBLISHED
04
AXIS OS pingtest.cgi SSRF via Missing validateaddr Call
tcptest.cgi calls the validateaddr binary before pinging anything. pingtest.cgi, accessible to viewer-level users, passes the ip parameter straight to busybox ping with no validation.
WRITEUP
2026
PUBLISHED
05
AXIS OS SNMP Community String Disclosure to Viewer-Privileged Users
The YAML access control definition grants viewer-level users GET access to SNMP community strings, while correctly restricting SET to admin only. Community strings are the sole authentication mechanism for SNMP v1/v2c.
WRITEUP
2026
PUBLISHED
06
Stack Overflow in JSONMergePatch Crashes Aiven Managed ClickHouse via Single SELECT Query
Any authenticated ClickHouse user, including SELECT-only accounts, can crash the entire managed instance in one HTTP request by passing two deeply nested JSON documents to JSONMergePatch, which recurses without a depth limit.
WRITEUP
2026
PUBLISHED
07
Authenticated DoS: Dragonfly Server Crash via Crafted Stream RESTORE Payload
A single RESTORE command with a crafted stream payload causes unbounded vector allocation in Dragonfly's RDB deserialization path, throwing std::bad_alloc and terminating the entire server process.
WRITEUP
2026
PUBLISHED
08
CVE-2024-32972: GetBlockHeaders Integer Underflow Causes Full Network Denial of Service
Sending Amount=0 in a GetBlockHeaders p2p message triggers a uint64 underflow to UINT64_MAX. The resulting allocation scales with chain height -- on Electroneum mainnet, a single connection is sufficient to OOM-kill the node.
WRITEUP
2026
PUBLISHED
09
Electroneum QBFT HasBadProposal Quorum Inconsistency Enables Permanent Consensus Stall
roundChangeSet.Add() trusts a single message's HasBadProposal flag to skip the digest check. isJustified() requires a quorum. One Byzantine validator can exploit this inconsistency to poison the prepared-block cache and stall consensus indefinitely.
WRITEUP
2026
PUBLISHED
10
User Email Enumeration via Error Message and Timing Difference at Aiven Login
The Aiven login endpoint returns distinct error messages and 10-35x response time differences for registered versus unregistered email addresses, enabling reliable unauthenticated account enumeration.
WRITEUP
2026
PUBLISHED
11
Persistent DoS via GZIP Compression Bomb in Aiven Karapace REST Proxy
An authenticated Kafka producer can crash the Karapace REST Proxy by writing a GZIP compression bomb to a topic, because librdkafka's GZIP codec has no decompression size limit while its ZSTD codec does.
WRITEUP
2026
PUBLISHED
12
Mattermost Shared Channel Invite API Missing Channel-Level Authorization
The invite and uninvite endpoints for shared channels enforce a system-level permission but skip the channel-level membership check. A user with manage_shared_channels can exfiltrate any private channel to a remote cluster they control.
WRITEUP
2026
PUBLISHED
13
Aiven Internal System Account Password Hashes Exposed via mysql.user SELECT
The avnadmin user on Aiven managed MySQL has unrestricted SELECT on mysql.user, exposing caching_sha2_password hashes for the internal root, replication, and monitoring accounts.
WRITEUP
2026
PUBLISHED
14
Root-Cause Analysis: OpenPGP.js CVE-2025-47934
A close read of the signature-verification bypass in OpenPGP.js v6: how a packet-list mutation during streaming verification allowed result.data to diverge from the verified content.
WRITEUP
2026
PUBLISHED
15
Autovacuum Arbitrary Code Execution via Expression Index Shadow Functions
Any avnadmin user can execute code in the PostgreSQL superuser context by chaining a gatekeeper bypass with expression index evaluation in the autovacuum background worker.
WRITEUP
2026
PUBLISHED
16
aiven_gatekeeper Bypass via Implicitly Castable Argument Types
Aiven's PostgreSQL security agent blocks shadow functions with identical signatures but misses those with different, implicitly castable argument types, defeating its own pghostile-class protection.
WRITEUP
2026
PUBLISHED
17
Incomplete CVE-2025-31480 Remediation: Unqualified parse_ident in SECURITY DEFINER Function
After fixing CVE-2025-31480 by schema-qualifying function calls across aiven_extras, one unqualified parse_ident call was missed in a SECURITY DEFINER body, leaving the same vulnerability class present.
WRITEUP
2026
PUBLISHED
18
Customer-Triggerable Superuser dblink Session via SECURITY DEFINER Chain
A SECURITY DEFINER function in aiven_extras opens a superuser database connection to localhost using trust authentication, reachable by any avnadmin user and from within logical replication apply worker triggers.
WRITEUP
2026
PUBLISHED
19
Privilege Boundary Violation via Subscription Ownership Escalation
The aiven_extras SECURITY DEFINER subscription creation function assigns ownership to the postgres superuser rather than the calling user, causing apply worker triggers to run with session_user=postgres.
WRITEUP
2026
PUBLISHED
20
Aiven Customer List Enumeration via 403/404 Response Differentiation on Project Endpoint
Any authenticated Aiven user can enumerate the platform's customer list by checking whether a project name returns 403 (exists, not a member) or 404 (does not exist), since project names frequently match company names.
WRITEUP
2026
PUBLISHED
21
Unauthenticated Query Storage and Cross-User IDOR in Aiven SQL Optimizer
The Aiven SQL query optimizer accepts queries without authentication and stores them so any authenticated user can retrieve another user's SQL by optimization ID, leaking table structures, column names, and business logic.
WRITEUP
2026
PUBLISHED
22
Two SSRF Vulnerabilities in a Cloud Image Pipeline
How inconsistent URL-validation coverage across four outbound-HTTP code paths produced two independent SSRF vectors in the same platform.
WRITEUP
2026
PUBLISHED
23
ASLR Bypass via Lua Function Pointer Leak in Aiven Managed Valkey
Any authenticated Valkey user can read 12+ C function pointer addresses from the server process by calling tostring() on Lua function objects, defeating ASLR for the valkey-server process.
WRITEUP
2026
PUBLISHED
24
Silent Data Corruption and Persistent Backdoor via Unrestricted redis.set_repl in Aiven Valkey
Any authenticated Valkey user can suppress replication of arbitrary write commands via Lua, silently diverging master and replica state, and can register persistent trojan functions that continue corrupting data after the attacker disconnects.
WRITEUP
2026
PUBLISHED
07

About

The short version.

I work at the seam between offensive and defensive security, and I build on both sides of it.

By day I'm a SOC analyst at a managed security provider, triaging 150–300 alerts a shift across Splunk, Microsoft Sentinel, and SentinelOne, where I cut false-positive escalations by roughly a third through detection tuning. Off the clock I run coordinated vulnerability disclosure on Bugcrowd and HackerOne and independent research with one hard rule: nothing is called a finding until it's reproduced.

The through-line is building. Detection-as-code pipelines validated by adversary emulation. A differential JWT fuzzer that turns library disagreements into auth-bypass findings. An autonomous vulnerability-research platform with a reproduce-before-report discipline. The work also reaches beyond security — a quant trading system, a computer-vision project, a shipped iOS app, and a live multiplayer game, each built from scratch.

I hold Security+ (DoD 8140 IAT II), SC-200, AZ-104, and AZ-900, with CySA+ scheduled. Based in Memphis and open to relocation. Every project here is something I designed, built, and ran — click in for the architecture, the code, and the evidence.

08

Résumé & Contact

The fastest ways to learn more and get in touch.

Let's talk.

Open to cybersecurity, security operations, analyst, research, technical security and support, and offensive/security-testing roles. Based in Memphis and open to relocation — email is the fastest way to reach me.