I build AI agents that do real work — and quantum models that describe real physics.
Ph.D. Candidate at UT Austin · Software Engineer · Vibe coder
By day I model quantum systems at the molecular scale. By night I ship AI-powered tools that solve problems I actually have — an agent that controls my Google Workspace, a tool that turns job descriptions into tailored CVs without sounding like AI wrote them, and open-source packages used by other researchers.
Currently open to software engineering and data science roles.
FlowSpace — Google Workspace, unified
An AI agent with 23 tools that reads, writes, and acts across your Drive, Gmail, Calendar, and Tasks. Not a chatbot — it actually sends emails, creates events, and finds documents. Write operations go through an approval step so you stay in control.
TypeScript Node Claude API Google APIs
kairos — Puts your potential into words
Most people undersell themselves not because they lack the work, but because they can't articulate it. kairos helps you surface what you've actually done, find the thread that connects it, and translate that into a tailored CV that reveals who you really are — not a polished version of someone else.
Claude Code HTML Prompt Engineering
| Project | What it does |
|---|---|
| Alprotein | Excitonic Hamiltonians and optical spectra for molecular aggregates |
| NeuralFinance | Neural networks for financial time-series prediction |
| pdbtoolkits | PDB protein structure parsing and analysis |
| publication-scripts | Reproducible code from my peer-reviewed papers |
Core
AI Engineering
The University of Texas at Austin — Ph.D. Candidate
Quantum mechanical modeling of molecular aggregates · optical spectra · excitonic systems
I write production-quality Python packages to model the physics of light-harvesting molecular systems, then validate them against experimental data. The code ships as open-source; the results go into journals.
→ Reproducible code from my papers
Want to collaborate or just talk shop? Reach out.