Portfolio
Miles
Ehrlich

CS & Math at Fordham · 3-2 Columbia Engineering · Builder & researcher building things that matter.

01 — About

Building things that actually matter.

I'm a Math & CS student at Fordham University with a 3.9 GPA, on a 3-2 track to earn a Computer Engineering degree from Columbia. I build full-stack products, lead ML research, and have a habit of turning hackathon weekends into real software.

Whether it's shipping a live AI education platform, winning Fordham's first hackathon, or digging into adversarial robustness of neural networks. I care about work that has real impact beyond the screen.

0
GPA at Fordham
0+
Projects shipped
0st
RamHack place
0
Research started

02 — Education

The journey.

2020 – 2024

Shaker High School

Advanced Regents Diploma with High Honors

2024 – 2027

Fordham University

BS, Mathematics & Computer Science

GPA 3.9 · CS Concentration · 3-2 Dual Degree Program

2027 – 2029 (expected)

Columbia University

BS, Computer Engineering

Via Fordham 3-2 Engineering Program

03 — Skills

What I work with.

Languages

PythonJavaScriptTypeScriptJavaC++SQLHTMLCSS

Frameworks

ReactNext.jsNode.jsExpress.jsTailwind CSS

Data & Infra

PostgreSQLREST APIsJWTOAuthGit / GitHubCanvas API

AI / ML

PyTorchCNN / Multitask LossAnthropic APIClaude VisionGoogle Gemini

04 — Projects

What I've built.

01Live

PromptlyPerfect

AI Education Platform

Founded an interactive platform that teaches users to write effective AI prompts through tiered XP/streak learning, a live scoring sandbox, and a community prompt library.

Full-stack: Node.js/Express, PostgreSQL, Anthropic API. Teams tier with role-based access and admin analytics.

Node.jsExpressPostgreSQLAnthropic APIJWTOAuth
Visit site →
02🥇 RamHack '24

EcoAudit

Sustainability Dashboard

Won First Place in the Environmental Track at Fordham's inaugural RamHack, a platform that identifies brand sustainability and carbon emissions data in real-time.

Canvas API perimeter-trace animation with Claude Vision. Leaflet.js + Clearbit APIs for an interactive global footprint map.

Vanilla JSCanvas APIClaude VisionLeaflet.jsClearbit
03Columbia DevFest

MINT

Browser Extension

A React browser extension that turns streaming media into an interactive shopping experience. Click and drag over any object in live video to identify and find it.

Google Gemini and Dedalus Labs APIs via Python backend for real-time image recognition and product matching.

ReactChrome ExtensionGoogle GeminiDedalus LabsPython
042025 – Present

Adversarial Robustness

ML Research · Fordham University

Leading undergraduate research on adversarial robustness of CNNs trained with human-supervised data augmentation, comparing baseline vs. human-supervised augmentation gates.

Multitask CNN with dual-objective loss function benchmarked across MNIST, CIFAR-10, and ImageNet.

PyTorchCNNMNISTCIFAR-10ImageNetPython

05 — Experience

Where I've worked.

Ostroff Associates Inc.

Albany, NY

Legislative AideJan 2024 – Aug 2025
InternJul 2023 – Jan 2024
  • ·Identified and secured government grant opportunities for clients, contributing to increased project funding.
  • ·Collaborated on research and analysis of legislative and policy developments to form strategic counsel.

Fordham University

New York, NY

Undergraduate Research Lead2025 – Present
  • ·Leading research on adversarial robustness of CNN models trained with human-supervised data augmentation.
  • ·Dual-objective loss function benchmarked across MNIST, CIFAR-10, and ImageNet.

CS Club, Fordham University

New York, NY

Secretary2024 – Present
  • ·Organizing technical workshops, events, and hackathon preparation sessions.

Leukemia Lymphoma Society

Albany, NY

Volunteer Fundraiser & Event Coordinator2022 – 2023
  • ·Raised over $15,000 in two months through strategic outreach to local businesses.

06 — Contact

Let's build
something.

© 2026 Miles EhrlichBuilt with Next.js & Framer Motion