https://github.com/dylan-sutton-chavez/dylan-sutton-chavez
My personal GitHub profile: introduction, activity and a snake eating my commits :).
https://github.com/dylan-sutton-chavez/dylan-sutton-chavez
github profile
Last synced: about 1 month ago
JSON representation
My personal GitHub profile: introduction, activity and a snake eating my commits :).
- Host: GitHub
- URL: https://github.com/dylan-sutton-chavez/dylan-sutton-chavez
- Owner: dylan-sutton-chavez
- Created: 2026-03-08T09:32:20.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2026-05-19T14:36:03.000Z (about 1 month ago)
- Last Synced: 2026-05-19T17:44:10.119Z (about 1 month ago)
- Topics: github, profile
- Homepage:
- Size: 45.9 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Dylan Sutton Chavez
**IEEE WCCI 2026 Peer Reviewer · ML Researcher · Systems Developer @ Amazon Audible**
---
Systems Developer at Amazon Audible, peer reviewer for IEEE WCCI 2026, and independent ML researcher with work currently under journal review.
I work on resource-constrained machine learning, compiler design, and production systems engineering. From migrating Player Services across AWS regions to building a single-pass SSA compiler in Rust, and proposing uncertainty-aware classifiers for TinyML deployment validated on NASA IMS and SemEval.
Most of what I build is small, fast, and deterministic. Open to collaborations on AI safety, selective classification, and embedded ML.
---
### Research
**Revisiting Rosenblatt Perceptron: Robust High-Entropy Classification via Uncertainty Margins**
First-author paper introducing an uncertainty-aware linear classifier with adaptive abstention margin for TinyML deployment. ~1 KB memory footprint, 9 ms latency, benchmarked against Bonsai, FastGRNN, ProtoNN, and LSTM.
- Paper source: [uncertainty-simple-perceptron.tex](https://github.com/dylan-sutton-chavez/uncertainty-simple-perceptron/blob/main/uncertainty-simple-perceptron.tex)
- Implementation: [uncertainty-simple-perceptron](https://github.com/dylan-sutton-chavez/uncertainty-simple-perceptron)
### Systems
**Edge Python** — Single-pass SSA compiler and threaded-code VM for a sandboxed Python subset. NaN-boxed values, dual inline caching, super-instruction fusion, pure-function memoization, mark-sweep GC; classes with inheritance and dunder protocol, async/await, pattern matching, and packages.json imports. around 170 KB WebAssembly module.
- Website: [edgepython.com](https://edgepython.com/)
- Live demo: [demo.edgepython.com](https://demo.edgepython.com/)
- Source: [edge-python](https://github.com/dylan-sutton-chavez/edge-python)
**Edge Python Official Packages:**
- Host Packages ([repository](https://github.com/dylan-sutton-chavez/edge-python-host)): Official packages that embed host-side bridge code (e.g., JS) and expose it to Python through the capability protocol. Includes modules like requests for networking and DOM bindings for browser interaction.
- Standard Packages ([repository](https://github.com/dylan-sutton-chavez/edge-python-std)): Official .wasm standard-library packages, where each capability is a Rust crate compiled to wasm32 against the wasm-pdk ABI. Hosts load the resulting .wasm over the standard plugin contract, no custom embedder, no Rust on the consumer side.
---
