Building production ML systems across computer vision, generative AI, and embedded deployment.
I hold M.S. degrees in Electrical Engineering & Computer Science and Behavioral & Computational Economics from Chapman University, and a B.S. in Business Economics from UC San Diego. My research and development interests include deep learning for image and video understanding, generative modeling, and Bayesian inference, with an emphasis on deployable solutions. I also instruct courses in Machine Learning and Unix/Linux Systems at the graduate and undergraduate levels.
- Generative models — GAN architectures with spectral normalization, attention mechanisms, and progressive training strategies
- Computer vision pipelines — U-Net segmentation, transfer learning (MobileNetV2, VGG16, ResNet), and video frame processing with temporal consistency
- LLM tooling — Prompt engineering, completion benchmarking, and AWS-based inference infrastructure
- Embedded ML — Low-latency systems on microcontrollers (nRF5340, ESP32) with real-time signal processing
- Statistical modeling — Bayesian GLMs in Stan/brms with posterior diagnostics, causal inference, and time series forecasting
| Project | Description |
|---|---|
| 🎨 GAN Image Colorization | ResNet U-Net generator + PatchGAN discriminator with spectral normalization and attention-augmented skip connections. 20% PSNR improvement over baseline. |
| 🖼️ U-Net Segmentation | MobileNetV2 encoder with 5-layer feature extraction and transposed convolutions. 87%+ validation accuracy on 7,400 images, optimized for edge deployment. |
| 🎥 Video Colorization | VGG16 autoencoder trained on 30K+ frames with Keras Tuner hyperparameter optimization. 95% temporal consistency. |
| 📡 BLE Audio System | nRF5340 + ESP32 broadcast system with isochronous channels, I2S routing, <150ms latency at 71ft range. |
| 📊 Bayesian Modeling | Stan/brms GLMs with weakly informative priors, posterior predictive checks, and 43% forecasting improvement. |
Architectures: GANs · U-Net · ResNet · VGG16 · MobileNetV2 · Transformers/ViTs
Techniques: Transfer Learning · Semantic Segmentation · Spectral Normalization · Attention Mechanisms · Bayesian Inference
- Google Advanced Data Analytics
- Google Cloud Fundamentals
- Microsoft C++ Programming
- AWS Certified Cloud Practitioner
Thanks for visiting! Always curious, always building. 👩🏻💻