From gravitational-wave simulations to production AI agents. I build things that decode complexity.
AI agents, astrophysics tools, hardware designs, and everything in between.
AI-powered missed call recovery and lead management platform for local businesses. Conversational voice agents handle inbound calls, extract lead data, and deliver structured summaries.
Persistent memory layer for AI agents. Uses Ebbinghaus-inspired decay curves to model memory salience over time. MCP server gives any LLM biologically-modeled recall, formation triggers, and age-aware retrieval.
Real-time orbital asset visualization platform with natural language search. NLP applied to astrophysical objects, search and filter satellites using intuitive commands. Tron/Encom-inspired aesthetic.
Full mechanical and interior restoration of a 1982 Datsun 280ZX. Fuel system rebuild, HVAC sensor diagnostics with multimeter tracing, vacuum line replacement, engine work, water pump overhaul, and custom audio install.
Gravitational wave background visualization from supermassive binary black hole systems as observed by Pulsar Timing Arrays. Interactive exploration of gravitational wave anisotropy across the sky.
Cosmological dark matter distribution simulation using ΛCDM parameters. Visualizes large-scale structure formation using the CLASS Boltzmann solver and Python scientific libraries.
Top-down galactic map of confirmed exoplanet locations using NASA's Exoplanet Archive API. NLP-powered search in progress for natural language querying of planetary properties.
Clean, responsive weather application with real-time data and forecasts. Location-based filtering with an intuitive interface. Built as a capstone CS project.
VGA signal generator built from scratch in SystemVerilog on an Intel DE10-Lite FPGA. Developed and tested in Quartus Prime and ModelSim, outputting display signals at standard resolutions.
I'm Peter Orndoff. A physicist turned builder. I work at the intersection of science and engineering, applying first-principles thinking to everything from AI agent systems to gravitational wave research.
I believe the best tools come from deeply understanding the problem. Whether it's modeling memory decay for an AI or simulating dark matter distributions, the throughline is the same: decode complexity, then build something useful from it.