I work at the intersection of neutron scattering science, AI research, and scientific software engineering — building the software that turns raw neutron events into physics, and exploring how AI can accelerate the way we do science.

• 🧲 Neutron scattering — data reduction and analysis software for neutron imaging and diffraction instruments at ORNL
• 🧠 AI research — applying foundation models and agentic systems to scientific workflows, from data reduction to automated analysis
• 🛠️ Software engineering — production scientific software in Python and C++, from detector event streams to user-facing analysis tools

• Most of my energy right now is going into NEREIDS, where I'm building out neutron transmission fitting features like a bounded multiplicative baseline model, physics-complete Ikeda-Carpenter calibration, and TIFF preprocessing.
• I'm hardening the imaging pipeline's robustness — tackling dead-pixel detection for intermittent/hot pixels and chasing down subtle issues like baseline E_ref being computed on the full grid instead of the active fit window.
• I've been refining fitter flexibility, adding per-parameter freezing and support for joint fit_energy_scale and fit_temperature fitting with better calibrate_energy performance.
• On the maintenance side, I'm keeping HyperCTui and timepix_geometry_correction secure by triaging Grype CVE findings for Python 3.12.

_{✨ This section is written by Claude (anthropic/claude-opus-4.8 via OpenRouter), which reviews my recent public GitHub activity on a schedule and summarizes what I've been working on. See how this profile works.}

Neutron scattering

• Mantid – Collaborative neutron and muon scattering analysis suite with global contributors.
• iMars3D – Neutron imaging reconstruction workflow powering instrument operations at ORNL.
• mcpevent2hist – Transforms raw MCP detector events into analysis-ready histograms.
• iBeatles – Utility collection that streamlines neutron imaging beamline experiments.

AI for science

• PLEIADES – LANL/ORNL research on scalable, AI-enhanced experimental workflows.
• NEREIDS – MCP-enabled neutron imaging analysis with new resolution models for AI-assisted workflows.

_{Last sync: Wednesday, July 8, 2026 at 6:19 AM CDT · 38 followers · 221 repos tracked (194 public) · 67 stars · Top languages: C++ (31.8%), Python (30.7%), HTML (25%), IGOR Pro (4.9%), Rust (2.4%), TypeScript (1.7%)}

_{🐍 regenerated daily from my contribution graph by Platane/snk}

This README is not edited by hand — it's the output of a small agentic pipeline that lives in this repo:

1. A scheduled GitHub Action runs scripts/update_profile.py via uv (zero-setup Python with PEP 723 inline dependencies) to pull live stats from the GitHub API and render this page from a template.
2. On a weekly cadence, scripts/ai_highlight.py feeds my recent public activity to Claude through OpenRouter (model-agnostic — one env var swaps the LLM) and lets it write the Currently Exploring section above.
3. The pipeline itself was built and is maintained with Claude Code — the same agentic tooling I use day-to-day for scientific software development.

_{Last updated on Wednesday, July 8, 2026 at 6:19 AM CDT.}