<Arman/>
OPEN TO WORK
GitHub LinkedIn Resume
CS STUDENT / ENGINEER/PORTFOLIO / 2026

Hi, I’m Arman Hassan.

Computer science + Data science · UNC ’27

I have a few ideas.
Fortunately, I know how to code.

Full stack development · LLM integration · Backend APIs · Databases · Cloud deployment

See my projects About me
+ A LOOK INSIDE THE SYSTEMSYS.001
01 / SOURCESCROLL TO EXPLORE
system / example trace
$ npm run build▌source → modules → application
SCROLL THROUGH MY PROJECTSCODE → SYSTEMS → PROJECTS
01MY PROJECTS

Selected projects.

What each project does,
and how it works.

Live demos available.

Jev 2048 live app with colorful boards, bot selection, seeded gameplay controls, and decision inspection.
/01Decision systems / interactive experiment

Jev 2048

Can a general-purpose AI classifier play 2048? Compare TypeSafe’s Jev model with a search algorithm and a pretrained reinforcement-learning bot, then inspect the information behind every move.

  • Pilot mean score 796 → 34,276
  • 67.1M-weight C++ model
  • 77,928 API requests for $5.45

My workGame engine, experiment design, native integration, and deployment

TypeSafe JevNext.jsReactTypeScriptExpectimaxTD learningC++Oracle CloudVercel
What I built Live Demo (opens in a new tab)Open the demo to play or compare bots. Jev uses a live API; service availability and credits can affect play.
CourseCompass AI homepage with its UNC catalog introduction, empty question field, and program exploration options.
/02RAG chatbot / UNC degree planning

CourseCompass AI

A degree-planning assistant for UNC students covering 200+ majors and minors plus general education. It answers from the official catalog with clickable citations and checks uploaded schedules against requirements.

  • Eval pass rate 11% → 100%
  • 200+ programs and tracks
  • 80+ offline tests in CI

My workRetrieval pipeline, guardrails, evals, UI, and deployment

Next.jsTypeScriptAWS BedrockClaudeTitan embeddingsS3 VectorsPythonGitHub ActionsVercel
What I built Live Demo (opens in a new tab)
Discord music illustration with headphones and musical notes; cover artwork, not a screenshot of the bot.
/08Discord music / backend systems

Arman MB

A free Discord music bot for friends, with live song suggestions, personal Spotify playlist connections, lyrics, and shared queue controls. Built in TypeScript and deployed on Oracle Cloud.

  • 19 slash commands
  • Up to 20 search suggestions
  • 104 automated tests

My workBot development, song matching, Discord UX, testing, and cloud deployment

TypeScriptNode.jsdiscord.jsLavalinkSpotify Web APIOracle Cloud
What I built Add to Discord (opens in a new tab)Free hosted bot; source code is private. Add it to a server you can manage, join voice, and use /play. Personal Spotify playlist login is subject to Spotify app-access and ownership rules.
VibeSafe website: Ship code fearlessly, with AI-powered security scanning.
/03AI security scanner agent

VibeSafe

VibeSafe scans repositories for security problems, posts findings on pull requests, and fails CI when it finds critical vulnerabilities. Built in collaboration with Humza Hassan.

  • 51 rules / 10 vulnerability categories
  • Fails CI on critical findings

My workSecurity scanner development

PythonClaude APIFastAPIDockerGitHub Actions
What I built Live Demo (opens in a new tab)
Industry Resilience Explorer website with industry drawdown, recovery, and resilience metrics.
/04Economic data / machine learning

Industry Resilience Predictor

An interactive dashboard and regression model that compare COVID-era drawdown and recovery across 90+ industries. Built during a hackathon.

  • 90+ industries compared

My workData pipeline, model & dashboard

Pythonpandasscikit-learnStreamlitDocker
What I built Live (opens in a new tab)
Other Projects
02EXPERIENCE

Where I’ve worked.

  1. Timing

    Software Engineering Intern

    January to May 2026

    • Built a GPT-4o agent to prioritize follow-ups and draft outreach across 1,000+ contacts. It combined function calling and structured outputs with a RAG pipeline using OpenAI embeddings, pgvector, and PostgreSQL.
    • Built a Manifest V3 Chrome extension at the early-stage AI startup, using content scripts and browser APIs to bring Timing’s networking assistant into LinkedIn and Gmail.
    • Reduced p95 contact API latency by 80%, from 450 ms to under 90 ms, with composite PostgreSQL indexes and Redis caching.

    TechnologiesOpenAI API · PostgreSQL · pgvector · Redis

  2. UNC Chapel Hill

    Teaching Assistant

    COMP 110 · Introduction to Programming

    August to December 2025

    • Helped 200+ students build their Python skills in weekly labs and office hours, working through bugs and concepts including control flow, functions, OOP, runtime analysis, and memory diagrams.
    • Graded programming assignments and gave specific feedback on students’ code, working with the course staff to address concepts students repeatedly found difficult.

    TechnologiesPython · VS Code

  3. Vogro

    Software Engineering Intern

    Stanford-affiliated

    May to August 2025

    • Helped build software for Vogro, a Stanford-affiliated volunteer program serving seniors. I worked with project leads on the platform and internal tools that supported the program’s day-to-day work.
    • Worked with project leads to plan and build REST API integrations using FastAPI and Node.js, connecting the React frontend, backend services, and third-party tools for 500+ users.
    • Wrote Python scripts to move, validate, and deduplicate data across internal tools, cutting time spent on manual data processing by 40%.
    • Improved page load times by 20% by removing duplicate API requests and refactoring React components, making the app run more smoothly on lower-end devices.

    TechnologiesPython · FastAPI · Node.js · React

03TOOLS I USE

What I
work with.

These are tools I’ve used in my projects and internships.

/01

Languages

  • Python
  • SQL
  • TypeScript / JavaScript
  • Go
  • Java
  • C++
  • HTML / CSS
/02

Frontend & Mobile

  • Next.js / React
  • Tailwind CSS
  • React Native / Expo
  • Playwright
  • Streamlit
  • Unity / Pygame
/03

Backend & Infrastructure

  • FastAPI / Node.js / REST APIs
  • PostgreSQL / Redis
  • Vercel / Oracle Cloud
  • AWS S3 / IAM
  • Docker / GitHub Actions
  • Git / GitHub
/04

AI & Data

  • OpenAI API / Claude
  • AWS Bedrock / Titan embeddings
  • RAG / pgvector
  • Tool calling / structured outputs
  • LLM evals / guardrails
  • Amazon S3 Vectors
  • scikit-learn / pandas / NumPy
Coursework at UNC Chapel Hill
  • Data Structures
  • Algorithms and Analysis
  • Systems
  • Computer Organization
  • Linear Algebra
  • Probability and Statistical Inference
  • Data Management
  • Machine Learning
  • Introduction to Artificial Intelligence
  • Mobile Computing Systems
  • Hardware Security
  • Models of Languages and Computation
  • Computational Photography
04ABOUT ME

0 / 27 characters

A bit
about me.

I’m a CS student at UNC Chapel Hill with a minor in Data Science, graduating in May 2027. Lately, a lot of what I work on comes back to AI. I like trying open source models and tools I find through Reddit or X, running them locally, and experimenting with them or turning them into small projects when something catches my attention. I like making things that make life easier for me and for others, or just genuinely interest me. I also enjoy working with other people, sharing ideas, and figuring things out together. Lately, I’ve been curious about distilling models and seeing how capable I can make smaller models run locally.

But right now, I'm more focused on learning Go and Elixir because I want to better understand concurrency and fault tolerance, and how those are used to build a backend infrastructure that scales properly.

UNC Chapel HillBS in Computer Science · Minor in Data ScienceGraduating May 2027
My resume
05GET IN TOUCH

HAVE A QUESTION OR AN IDEA?

Want to

I’d be glad to hear from you.

UNC Chapel Hill
GitHub (opens in a new tab)LinkedIn (opens in a new tab)Resume