Brandon Davidson
Founding Engineer @ iya | Software, AI & Data Engineering
Atlanta, Georgia
Computer Science undergraduate at Georgia Tech building full-stack agronomy systems, machine learning pipelines, and agent-ready developer tools.
Projects
- Brandon Brain- Rust, SQLite, Tokio, Clap, Obsidian, Docling
Built an Obsidian-first knowledge compiler with CLI workflows for workspace initialization, health checks, ingest queues, and OCR bootstrap.
- Implemented SQLite state, file walking, hashing, config persistence, and Markdown/PDF ingestion for agent-ready artifacts.
- Packaged install and doctor commands to validate local runtimes, workspace paths, and OCR dependencies.
- Biomechanical Stress Analysis- Python, OpenCV, YOLOv8, scikit-learn, PyTorch
Built an end-to-end computer vision system to predict biomechanical injury risk from short exercise videos.
- Extracted pose keypoints with YOLOv8 and engineered joint-angle, velocity, and movement-stability features.
- Trained Random Forest and neural network models to predict continuous stress scores with 84% accuracy.
Experience
- Founding Engineer- iya
Building a full-stack agronomy platform for field records, weather, soil, satellite, and crop-stage decision support.
- Developed a Python PCSE/WOFOST agronomy engine with Gymnasium RL training, baselines, and trace export.
- Implemented offline-first pipelines for NASA POWER weather, soil caches, Sentinel metadata, and vegetation features.
- Shipped React/Vite product surfaces and discovery workflows for scouting, alerts, records, and agronomic explanations.
- Built tests and playback tooling to compare learned policies against heuristic agronomy baselines.
- Software Engineer (AI & Machine Learning)- GROWER Lab
Built a React.js geospatial dashboard for real-time outage analysis across 50+ U.S. states.
- Developed Python data pipelines ingesting records from 1,000+ utilities for national resilience monitoring.
- Designed and evaluated machine learning forecasting models, reaching 85% accuracy on historical outage data.
- Analyzed millions of outage records with pandas, matplotlib, and Seaborn for DOE-facing reports.
- Data Engineering Intern (Technology Consulting)- Protiviti
Automated Azure ETL pipelines processing 10K+ daily rows for production business intelligence systems.
- Migrated 10 complex SQL stored procedures to Microsoft Fabric, reducing reporting runtime by 40%.
- Designed 5 production SQL views integrating tariff metrics into SAP-backed reporting workflows.
- Presented bi-weekly technical updates to align data models with client reporting requirements.