CS @ Georgia Tech · Intelligence (AI/ML) and Modeling and Simulation
Building autonomous systems from real and simulated data.
I love building at the intersection where models meet the physical world, because I think computer science should complement our day-to-day lives, not replace them. Currently building Tycho, a deterministic verification layer for AI agents, and working on a perception model for Georgia Tech's HyTech Racing team.
|
A file manager that launches Claude Code from any folder in one click. Sessions open as tabs inside the app and live in named spaces you switch between, so working on several projects at once stays seamless. Tabs and spaces are saved across restarts, and an "open recent" list pulls back recently opened Claude sessions. A built-in file explorer and terminal, plus one-click opening in your IDE, keep your development tools organized in one place.
|
Electron, React and TypeScript, with real ConPTY terminals, a loopback MCP server that lets a session drive the app itself, bundled ripgrep search, and undo on every file operation.
|
A grouped app tray and window switcher for Windows and macOS. One key shows every app you care about, sorted into groups you set up once, with open apps lit and closed ones dimmed. Jump to one or launch it with a click, an arrow key or a few typed letters, and see unread pings from your chat apps on a slim side panel without opening anything. Instead of scanning to see where Alt+Tab is taking you, you go straight to the window you have in mind, like macOS Spotlight with extra features.
|
Electron and JavaScript with a native Swift helper on macOS, an Alt+Tab replacement with type-to-launch search, a music strip with global media keys, Open All / Close All, optional on-device voice commands, and per-monitor layout that handles mixed scaling.
Languages
Machine learning & AI
Web & app
Tools & infrastructure
| Project | What it is | Stack |
|---|---|---|
|
Checks whether an AI agent's tool calls actually did what the agent claimed. The verdict engine runs offline with no model in the loop, reasoning over git state, filesystem diffs, exit codes and database deltas. Open source for Claude Code, Cursor, Codex and OpenCode, and took it through a Y Combinator interview. | Python · LLM integration · PyPI/npm | |
|
HEARSAY |
Detects whether a voice clip is real human speech or AI-generated (text-to-speech or voice cloning). Fuses three fine-tuned XLS-R speech models with speech-biology features, trained on 190,000+ clips from 11 open corpora plus fakes from our own synthesis simulator. Our initial submission hit a minDCF of 0.0584 and a 2.5% EER on the official test set. | Python · PyTorch · XLS-R · LightGBM |
|
Physics-informed NN on HTEM DB |
A year of mentored research on NREL's materials database. Built a ~9,500 sample by ~1,300 feature training set, then benchmarked 20 deep networks against a scikit-learn baseline for a 96.59% drop in MSE. Adding physics equations as learning constraints took the best architecture to an R² of 0.9932. | Python · PyTorch · scikit-learn · Jupyter |
|
Multimodal GNN for semiconductors |
Predicts band gap and formation energy straight from crystal structure across 17,000+ stable semiconductors from the Materials Project. R² of 0.81 on band gap and 0.98 on formation energy, with a custom PyTorch C++ extension that made CPU training ~2.5× faster. | Python · C++ · PyTorch Geometric · pymatgen |
Also contributed to: Pebble · Parky
Georgia Institute of Technology · B.S. Computer Science, 2029 · Atlanta, GA
LinkedIn ·
dmano6@gatech.edu