I’m a software engineer and researcher passionate about building intelligent, user-centered solutions. With a background in full-stack development and applied machine learning, I enjoy exploring how artificial intelligence can be leveraged to solve real-world problems.
At Stoke Space I build manufacturing software on Boltline. My work also spans designing iOS applications for accessibility, developing scalable IoT systems on AWS, and building AI models for computer vision and natural data classification.
I thrive at the intersection of software engineering and AI: translating complex data into actionable, meaningful applications. Whether it’s deploying cloud-based systems, training neural networks, or prototyping accessible mobile apps, I’m driven by curiosity and a commitment to developing solutions that make an impact.
Work Experience
- Led a year-long cutover of inventory auditing onto a single activity feed: closed logging gaps from 76% to 100%, backfilled ~1.75 million historical records with an idempotent dedupe Temporal pipeline, shipped who/when/where/what filters across 10+ surfaces, and removed dual writes from the hot path.
- Built an evaluation harness for Claude-based matching of historical inventory events to existing activity rows, iterating prompts against labeled cases until match quality held while cutting token use by 37%.
- Owned automatic percent complete for hardware build tasks. Cut Temporal fanout from O(N) to O(1) per work item after instrumenting the bottleneck in Datadog.
- Sole engineer for customer-defined inventory flags. Users create custom issue tags and automation rules instead of single hardcoded defect tags, then see those tags on inventory in the build tracker.
- Led 0-1 for manufacturing task scheduling. Turned a vague schedule ask into one model with a DAG and a Gantt, evaluated OSS graph and Gantt libraries vs building in-house, and built the end-to-end implementation.
- Spearheaded and owned the development, deployment, and operation of a coffee chat bot that automatically pairs employees for weekly 1:1s. Scaled adoption to a company-wide 500+ users, plus additional sub-teams concurrently.
- Directed the end-to-end design and prototyping of an iOS communication app in Swift for children with speech impediments.
- Conducted user requirement gathering with PMs and Speech Pathologists to design the back-end SQLite database to support personalized communication workflows for various use-cases.
- Led the research and evaluation process to select a secure remote development solution for collaborative code access. Established criteria around security, speed, cost, and accessibility.
- Led the migration of 3 primary codebases to secure, cloud-based version control on AWS, improving development workflow for 5 engineers working with the codebase.
- Designed and implemented an LSTM based AI model to predict radio wave behavior.
- Performed and documented black box testing to validate system functionality and support reliable software delivery.
- Debugged and maintained critical legacy code in C, contributing to stability and performance improvements.
Projects
ClosedSnow
A snow-forecasting web app for four Washington ski resorts, combining live weather, ops status, and avalanche danger in one dashboard.
AI Fungi Identifier
Enchancing Mushroom Safety through Backprogated Multi-Layer Perceptron Neural Network Identification
IoT Water Monitor
End-to-end IoT system tracking rainwater levels in outdoor storage tanks to reduce household utility costs.
AI Vehicle Detection
A comparative study of deep learning models for vehicle detection and segmentation, analyzing trade-offs between accuracy and inference speed.
Inverse Assembler
A command-line disassembler built from scratch in 68K Assembly to translate machine code into human-readable instructions.
Image Deep Learning
A Convolutional Neural Network in MATLAB for ASL letter classification, achieving 2nd place in a Kaggle competition.
Skills
C
C++
Python
Java
Swift
SQL
Cuda
Bash
ARM Assembly
Motorola 68000 Assembly
MATLAB
Arduino
JavaScript
TypeScript
HTML
CSS
RESTful APIs
Svelte
Vite
Flask
React
Next.js
Node.js
Tailwind CSS- Bootstrap
JSON
AWS
Azure
Vercel
Docker
Git
GitHub Actions
Temporal
DataDog
MySQL
PostgreSQL
DynamoDB
Supabase
TensorFlow
PyTorch
Keras
scikit-learn
Pandas
NumPy
NLTK
Seaborn
Matplotlib
Google Gemini API
Claude API
Visual Studio Code
Figma
Qualtrics
ArcGIS Pro
LaTeX
Education
