AT

Hi, my name is

Anthony Tang.

Ibuildlarge-scaleAIandwebsystems.

Student at Rice University (BS CS, BA Statistics, Minor Data Science). Full-stack & AI/ML. Prev @ WhyAI. Co-authored AAAI-26 paper on interpretability and LLM failures.

Anthony Tang

About Me

I was born in Round Rock and grew up in Austin, TX. I've always been curious how things work and loved building things of my own, which led me to CS, where I could combine problem-solving, creativity, and curiosity to build systems that help people. Outside of code, I enjoy reading, tennis, pickleball, and time with my dog! I played trombone and bass trombone for seven years, which taught me teamwork, discipline, and communication. Those experiences shape how I work: clarity, collaboration, and ownership. I'm a fast learner and like taking on complex, large-scale systems, whether refactoring backends, real-time translation for wearables, or researching why language models fail.

Education

Rice University

May 2027 · Houston, TX

BS Computer Science, BA Statistics, Minor Data Science · GPA 3.85/4.0

Coursework: Computational Thinking, Data Structures & Algorithms, Program Design, Computer Systems, Programming Languages, Concurrent Programming, Computer Security, Data Science, Linear Regression, Probability & Statistics, Multivariate Calculus, Linear Algebra

Where I've Worked

Software Engineer Intern

Oct 2025 – Jan 2026

WhyAI · Houston, TX

  • Spearheaded large-scale backend refactor: migrated 70%+ of codebase to layered microservices (SOLID), eliminating technical debt.
  • Built cloud-native CI/CD with Pytest & Playwright; 80% coverage, 99.9% deployment reliability.

Software Engineer Intern

May 2025 – Aug 2025

Curriculum Associates · Boston, MA

  • Optimized data viz for 500+ developers: high-performance query for 100k+ points, 50% productivity gain.
  • Led React.js migration of mission-critical asset management; improved responsiveness and maintainability.
  • 1st place hackathon: AI transcript generator (NLP), 50% faster content creation.

Software Engineer Intern

May 2024 – Aug 2024

Curriculum Associates · Boston, MA

  • Agile work on enterprise CMS: API optimizations and critical bug fixes for 500+ DAU.
  • Jira automation scripts removed 5+ hrs manual work per sprint.
  • 1st place hackathon: third-party SDK integration, 10% team efficiency gain.

Some Things I've Built

Research · AAAI-26

Why It Failed: A Benchmark to Evaluate Interpretability

LM4UC Workshop, AAAI-26. Co-authored with Joel Mathew, Aditya Lagu, Prudhviraj Naidu (Algoverse AI Research / UC San Diego). Benchmark for whether interpretability methods explain model failures. Last-token logistic probes on Gemma-2 2B across four reasoning tasks achieve near-chance performance, highlighting a gap for AI safety and failure prediction. We provide a standardized framework and argue for explaining failure modes, not just reporting metrics.

Python | PyTorch | Scikit-learn | Gemma-2 2B

Featured Project

Adapt Language Learner

Real-time full-stack translation for wearable tech (AugmentOS SDK). WebSocket streams tuned for sub-second latency in live captioning. Server-side pipeline with Supabase + regex for automated feature extraction (idioms, conversation metrics).

TypeScript | React | Node.js | Supabase | PostgreSQL

Other Projects

OwlMart

Student marketplace for the Rice University community

Next.js | TypeScript | Supabase

Job Monitor

Add job page URLs and run a check to see which links have new listings; view job titles at a glance.

Next.js | Playwright | TypeScript | Supabase

Roommate Matcher

Rice University app using weighted cosine similarity to match students for roommates and housing.

Next.js | GraphQL | Prisma | Celery | Redis

OwlTrack

HackRice 14 Adobe Challenge Winner. AI-powered 4-year course planner and course suggestions from vector distance.

Next.js | TypeScript | Python

Aetherborn

2D party fighting game built with GDScript and the Godot game engine.

GDScript | Godot

Get In Touch

Open to new opportunities, research, and building things together. Reach out via email, phone, or LinkedIn. I would love to chat!

LinkedIn

Austin, TX