Large-scale data analysis
Worked with TB-scale experimental datasets, simulation campaigns, and statistically rigorous validation pipelines to extract reliable results from complex data.
About me
I have a PhD in experimental particle physics and several years of experience building analysis pipelines, running large-scale compute workflows, and extracting reliable results from messy real-world data.
My work sits at the intersection of large-scale data analysis, research-driven scientific computing, and practical AI development. I build reproducible workflows with Python and C++, work comfortably in Linux and HPC environments, and communicate technical results through publications, talks, and project documentation.
Worked with TB-scale experimental datasets, simulation campaigns, and statistically rigorous validation pipelines to extract reliable results from complex data.
Led particle-physics analyses from method design to publication, combining statistical modeling, scientific software, and clear communication for technical audiences.
Built practical ML applications using embeddings, retrieval-augmented generation, and LLM APIs, with a focus on useful tools rather than demos.
Selected work
Python · Multithreading · Automation
Built a multithreaded command-line tool that automates SDME and PWA workflows, reducing repetitive manual work across large GlueX analyses.
RAG · Search · Embeddings
Retrieval-augmented search system combining vector embeddings with traditional keyword scoring for fast, AI-assisted document lookup.
LLMs · API calling · CLI tooling
Command-line translator supporting local Hugging Face models and remote LLM providers including OpenAI, Anthropic, and Gemini.
Frontend fundamentals
A collection of small web projects used to strengthen HTML, CSS, and JavaScript fundamentals while improving UI implementation skills.
Published research
Led a spin-density matrix element measurement of photoproduced φ(1020) mesons decaying to KSKL, using the full GlueX-I dataset. These results were published in Physical Review C.
Working paper
Compared competing fit models, validated a persistent two-peak structure, and documented the analysis in a working paper built from the full GlueX-I dataset.
Core toolkit
Python, C++, Go, SQL, JavaScript, HTML, CSS
Statistical modeling, maximum-likelihood fitting, uncertainty estimation, simulation validation, data visualization
Linux, Git, HPC environments, HTCondor, swif2, command-line tooling, workflow automation
RAG pipelines, vector embeddings, keyword retrieval, LLM API integrations, local model workflows
Academic foundation
Research output
Communication
Pedagogy