Matt Scruggs

About Matt Scruggs

I build and own highly available + scalable services, APIs, UIs and prototypes, and predictive modeling applications that redefine what’s possible.🧠 My experience includes machine learning (model research and development in NLP, LLMs, and computer vision) and other domains such as GIS and full-stack development. My superpower is learning how to solve any problem through curiosity and persistence; time and determination are the only factors. I am a fast, motivated, and independent learner (self-taught in many aspects of software), and enjoy breaking tough problems into smaller ones. I also enjoy finding and fixing issues before they affect customers, and pinning down and resolving unexpected performance issues.Some of my core philosophies:* compassion* quality over quantity* be direct* measure, measure, measure* do it right the first time* know what you don\'t know* ask questions* learn from mistakes* computers are tools; people matter

Experience

1. Staff Software Engineer

   KoBold Metals

   Apr 2022 — Present

   Collaborating with world-renowned geologists and data scientists to find critical battery mineral resources, with a focus on data engineering and making data more integrated and accessible to internal exploration teams.Major software projects I\'ve lead or contributed heavily to (with full-stack development across the board)- A scalable geospatial data platform (with a UI and API) built with open-source tools such as PostgreSQL, GDAL, QGIS, Django, Django REST Framework, Pydantic, Leaflet, and GeoPandas; leverages cloud-native raster and vector data formats for high-performance data access- A mineral exploration data taxonomy system for enriched data comprehension and discovery- A scalable document processing pipeline and full-text search search system for a wide variety of file formats, built with pub/sub components and OpenSearch- Rapid internal app development using Streamlit and Jupyter notebooks, and related research projects for AI-enhanced mineral exploration- Experiments with multimodal LLMs for delivering geological insights, from images and text.I also mentor and lead software teams, and help junior engineers develop their engineering skillsets while building systems like those described above.

Education

• The University of North Carolina at Chapel Hill

  BS, Environmental Science

  2007 — 2011

Skills

• Environmental Science
• Apache Spark
• Object Oriented Design
• Energy
• Numpy
• Java
• Mongodb
• Relational Databases
• Nosql
• Python
• Parallel Computing
• Genetic Algorithms
• Climate Change
• Rest
• Programming
• Html
• Gis
• Renewable Energy
• Django
• Sql
• Restful Webservices
• Javascript
• Adobe Livecycle Designer
• Css