Software Engineer - Machine Learning
John Hartquist
Full-stack software engineer who takes ideas from prototype to production, with expertise in ML, Python, TypeScript, and Rust.
Currently working on training real-time polyphonic pitch detection models in Python, and deploying them in Rust.
While some of the code for this website was generated with AI, the text is 100% human written.
Open to full-time and contract roles →Latest Writing
Resonators: A Rust Port of the Resonate Algorithm
Real-time spectral analysis without FFTs, ported to Rust with Python and WebAssembly bindings.
Apr 29, 2026
Fine-Tuning ResNet-18 for Audio Classification
Sweeping spectrogram hyperparameters with W&B to fine-tune ResNet-18 and DenseNet-161 on the ESC-50 audio classification dataset.
Oct 23, 2020
Audio Classification using FastAI and On-the-Fly Frequency Transforms
An experiment with generating spectrograms from raw audio at training time with PyTorch and fastai v1.
Nov 28, 2018
Projects
resonators
Rust implementation of the Resonate algorithm for spectral analysis with Python and TypeScript/WASM bindings.
candle-pitch
Porting pitch detection models to Rust via Hugging Face's candle framework.
claude-remotion-kickstart
Create videos programmatically with Claude Code and Remotion.
How I Vibe Code Technical Videos With Claude Code and Remotion
Video walkthrough of the claude-remotion-kickstart workflow.
Real-Time Audio Visualization in the Terminal
Rust TUI rendering raw waveform and FFT of microphone input in real time, built with ratatui and cpal. Work in progress.
cubed
A 2×2 Rubik's Cube with Rust/WASM move logic and React Three Fiber 3D rendering.
Real-time Guitar Analysis and Tab Transcription
Real-time polyphonic guitar transcription. Demo for my Master's Thesis in 2012.