zsh · ammaar@portfolio

Last login: today on ttys001

ammaar@portfolio ~ % neofetch

[ click the orb · warp to pi's island ]

Hey, visitor

Ammaar Rehman

Profile

ROLE
Cyber Security Analyst // Cyber Ready Clinic (intern cohort)
ROLE
Cloud & AI Operations Intern // NextGen Consulting
EDU
HCC A.A. (Dec 2026) → UMD Smith, B.S. IS (Jan 2027)
LOC
Ellicott City, MD

Certs

AWS
Cloud Practitioner certified
COMPTIA
Tech+ certified
READY5
Practitioner in cohort

Homelab

V2
one Pi 5, built with Ansible

Good evening, thanks for stopping by.

ammaar@portfolio ~ % open resume.pdf gitlab linkedin mail

about

Ammaar Rehman
ellicott city, md

I'm finishing my A.A. at Howard Community College this December on a 3.65 GPA, Dean's List, and I start at the University of Maryland's Smith School of Business in January 2027 to study Information Systems.

This fall I joined Cyber Ready Clinic's analyst cohort. Working alongside senior practitioners, I support cyber-readiness assessments for nonprofit clients using the Discover5 and Ready5 frameworks, with findings aligned to the NIST Cybersecurity Framework and CIS Controls.

At NextGen Consulting I started as an IT Operations Intern and closed out 200+ user issues. I'm now a Cloud & AI Operations Intern on the same team, using Claude and AWS Kiro to investigate errors and triage tickets, and to improve our support documentation.

Outside work I build and test things in the open. My first homelab ran high-availability DNS on two Raspberry Pi 5 nodes behind a keepalived floating IP, with failover tested in both directions. I'm now rebuilding it as one Pi, set up from Git with Ansible. In trading-lab I wrote the pass/fail rules for nine trading experiments before touching any data, and rejected every one before a dollar moved.

certs
AWS Cloud Practitioner (CLF-C02) · CompTIA Tech+
learning
Anthropic AI Fluency & Claude developer courses

skills

What I work with: certified, shipped, or currently being drilled.

ammaar@portfolio ~ % ls ./cloud_infra

experience

SEP 2026 – PRESENT
Maryland · internship cohort

Cyber Ready Clinic

Cyber Security Analyst

  • Support nonprofit cyber-readiness assessments using the Discover5 business-risk and Ready5 capability frameworks, with findings aligned to the NIST Cybersecurity Framework and CIS Controls
  • Analyze and document capability gaps across endpoint, identity, email, edge, and response controls, including EDR/MDR, MFA and Conditional Access, SPF/DKIM/DMARC, DNS filtering, and backup and restore testing
  • Participate in client engagements alongside senior practitioners, including leadership intake interviews, capability discussions with client IT teams and MSPs, and close-out briefings
  • Working toward the Ready5 Practitioner digital badge, expected Nov 2026 on cohort completion
MAY 2025 – PRESENT
DC–Baltimore area

NextGen Consulting Inc.

Cloud & AI Operations Intern

  • Use Claude, AWS Kiro and AWS services to investigate errors, summarize issues, and improve documentation
  • Run configuration checks, service testing, and issue triage on active cloud projects
  • Translate technical problems into clear next steps for non-technical users and stakeholders

Information Technology Operations Intern · Jun 2024 – May 2025

  • Resolved 200+ user issues spanning account configuration, access, and mobile app troubleshooting
  • Tested app functionality, documented bugs, and built support documentation to reduce repeat-issue resolution time
AUG 2024 – PRESENT
Columbia, MD

HCC Campus Store · Slingshot

Campus Operations Associate

  • Operate order, fulfillment, and inventory systems serving students, faculty, staff, and departments in a fast-paced, high-volume environment

projects

Four featured builds, then the rest.

A dig query to the homelab DNS returning six answers for google.com, with the server address redacted
RETIREDhomelab / networking

High-availability DNS

Version one of my homelab, retired September 2026. Two Raspberry Pi 5 nodes ran AdGuard Home to filter DNS for my own devices, behind a keepalived VRRP floating IP. keepalived moved the IP to the backup node automatically; failover and failback tested in both directions, config synced every 10 minutes. Version two, a single Pi built with Ansible, is in progress.

High-availability DNS, homelab v1, retired September 2026 Two Raspberry Pi 5 nodes ran AdGuard Home to filter DNS for my own devices, behind a keepalived VRRP floating IP. keepalived moved the IP to the backup node automatically; failover and failback tested in both directions, config synced every 10 minutes. Version two, a single Pi built with Ansible, is in progress. Homelab v1 · retired September 2026 my devices floating IP keepalived VRRP Raspberry Pi 5 / AdGuard Home node Raspberry Pi 5 / AdGuard Home backup node IP moves on failure config sync every 10 min failover and failback tested both ways Version two, a single Pi built with Ansible, is in progress. High-availability DNS, homelab v1, retired September 2026 Two Raspberry Pi 5 nodes ran AdGuard Home to filter DNS for my own devices, behind a keepalived VRRP floating IP. keepalived moved the IP to the backup node automatically; failover and failback tested in both directions, config synced every 10 minutes. Version two, a single Pi built with Ansible, is in progress. Homelab v1 · retired September 2026 my devices floating IP keepalived VRRP Raspberry Pi 5 / AdGuard Home node Raspberry Pi 5 / AdGuard Home backup node IP moves on failure config sync every 10 min failover and failback tested both ways Version two, a single Pi built with Ansible, is in progress.
Raspberry Pi 5 ×2keepalivedAdGuard HomeDocker
repo ↗
trading-lab's saved result for experiment 9, a trend-following allocation: four pre-registered checks pass, the 50 bp cost check fails, verdict REJECT
DONEresearch / python

trading-lab

Do popular trading strategies actually have an edge? Nine experiments: stock mean reversion, a crypto trend filter, Kalshi 15-minute markets, dual momentum, copying Congress, a holiday effect, a Kalshi model search, buying the close and selling the open, and a published trend-following allocation. Pass/fail rules committed to git before any data, backtests built to catch their own bugs, each judged on a single run. All nine rejected, $0 lost.

trading-lab: how each experiment was tested Nine experiments. Pass/fail rules committed to git before any data, backtests built to catch their own bugs, each judged on a single run. All nine rejected, $0 lost. nine experiments pass/fail rules committed to git before any data backtests built to catch their own bugs each judged on a single run all nine rejected · $0 lost trading-lab: how each experiment was tested Nine experiments. Pass/fail rules committed to git before any data, backtests built to catch their own bugs, each judged on a single run. All nine rejected, $0 lost. nine experiments pass/fail rules committed to git before any data backtests built to catch their own bugs each judged on a single run all nine rejected · $0 lost
Pythonpre-registrationbacktestingCI self-tests
repo ↗
Eval harness output: prompt v2 scores 15 of 15 on primary accuracy and 9 of 10 on multi-need coverage
SHIPPEDai / automation

Claude intake triage assistant

Turns messy free-text help requests into structured, reviewable records (category, urgency, summary, next step, human-review flag) via Google Forms, Apps Script, and the Claude API. A 15-case eval harness took classification accuracy to 100% and multi-need coverage from 0% to 90%. Every record carries a human-review flag.

Claude intake triage assistant: how a request becomes a record Turns messy free-text help requests into structured, reviewable records (category, urgency, summary, next step, human-review flag) via Google Forms, Google Sheets, Apps Script, and the Claude API. A 15-case eval harness took classification accuracy to 100% and multi-need coverage from 0% to 90%. Every record carries a human-review flag. Google Forms free-text help request Google Sheets Apps Script Claude API structured, reviewable record category urgency summary next step human-review flag every record carries a human-review flag 15 eval cases Python evals classification accuracy to 100% multi-need coverage 0% → 90% Claude intake triage assistant: how a request becomes a record Turns messy free-text help requests into structured, reviewable records (category, urgency, summary, next step, human-review flag) via Google Forms, Google Sheets, Apps Script, and the Claude API. A 15-case eval harness took classification accuracy to 100% and multi-need coverage from 0% to 90%. Every record carries a human-review flag. Google Forms free-text help request Google Sheets Apps Script Claude API structured, reviewable record category urgency summary next step human-review flag every record carries a human-review flag 15 eval cases Python evals classification accuracy to 100% multi-need coverage 0% → 90%
Claude APIApps ScriptGoogle SheetsPython evals
repo ↗demo ↗
Historical deployed static site showing homelab projects and a footer noting it was served from a private S3 bucket through CloudFront and provisioned with Terraform. The repository link is redacted.
SHIPPEDcloud / IaC

AWS static site with Terraform

A private S3 bucket served over HTTPS through CloudFront with Origin Access Control, provisioned entirely with Terraform. Deployment and teardown are scripted and documented.

AWS static site with Terraform: architecture A private S3 bucket served over HTTPS through CloudFront with Origin Access Control, provisioned entirely with Terraform. Deployment and teardown are scripted and documented. Terraform provisioned with Terraform visitor CloudFront Origin Access Control (OAC) private S3 bucket HTTPS access via OAC deploy and teardown scripted and documented AWS static site with Terraform: architecture A private S3 bucket served over HTTPS through CloudFront with Origin Access Control, provisioned entirely with Terraform. Deployment and teardown are scripted and documented. Terraform provisioned with Terraform visitor HTTPS CloudFront Origin Access Control (OAC) access via OAC private S3 bucket deploy and teardown scripted and documented
TerraformS3CloudFrontOAC
repo ↗
dns-filtering/ AdGuard Home on a Raspberry Pi, filtering DNS for my own devices. Where the homelab started. V1 monitoring-stack/ Prometheus scraped host and container metrics and Grafana charted them, in Docker Compose. V1 uptime-monitoring/ Uptime Kuma in Docker tracked availability and response health for the homelab's services. V1 remote-access/ A Tailscale mesh VPN (WireGuard underneath) gave SSH and admin access from anywhere, with no router ports opened. V1 ammaarrehman.github.io/ This site. Plain HTML, CSS, JavaScript and a little Python, no frameworks. The orb, the island, and the boot screen are all canvas. ACTIVE

education & certs

Howard Community College

DEGREE
A.A. General Studies
DATES
Aug 2024 – Dec 2026 CURRENT
GPA
3.65 · Dean's List
COURSES
Accounting I & II, Macro & Micro Economics, Business Calculus, Statistics, Marketing, Intro to Business, Computer Concepts, Python

UMD · Smith School of Business

DEGREE
B.S. Information Systems
DATES
Starts Jan 2027 · expected 2028 INCOMING

Certifications

AWS
Certified Cloud Practitioner (CLF-C02) Jun 2026
COMPTIA
Tech+ Aug 2026
READY5
Practitioner, Cyber Ready Clinic in progress · expected Nov 2026
ANTHROPIC
AI Fluency & Claude developer courses Jun 2026

Honors

HCC
Dean's List

contact

Want to talk? Email is the fastest way to reach me.

ammaar@portfolio ~ % mail ammaarrehmanf@gmail.com