Felix Dulys

About Felix Dulys

I thrive in dynamic startup environments where resourcefulness is key, but I also know how to navigate and deliver in larger organizations. My philosophy is simple: use the right tool for the job. If a spreadsheet is the most effective solution, I use it. If the solution doesn\'t exist, I have the technical depth to invent it—a mindset that led to my 2022 patent for a blockchain search algorithm.I view AI and LLMs as powerful force multipliers that improve my efficiency, but they are not crutches. I am an engineer first. I understand the limitations of current AI models, and I pride myself on the ability to teach myself new languages, dig into the documentation, and do the heavy lifting manually when required.My background spans the full lifecycle of machine learning and data science. I’ve built production-level applications including a blockchain scoring system (pyspark, GraphFrames), a home energy report calculator (keras), recommendation systems (python, scikit-learn), and a workflow to create 3D materials with multiple AI models (VertexAI with nano banana and gemini pro, Adobe Substance). I\'ve refactored and rebuilt an electronic vehicle detection algorithm based on home energy data (pyspark, airflow), a kitchen layout engine based on LiDAR and enterprise data (python, blender, node.js). Whether I\'m designing scalable data architectures on AWS/Databricks or making an Excel spreadsheet to work cross-functionally, my goal is to leave every project and team better than I found it.

Experience

1. Principal Engineer, Data Science

   MasterBrand, Inc.

   Jan 2025 — Present

   Hired & promoted from BroadPoint Group contract- Data architecture, science & engineering for full-stack application: Refactored and rebuilt a novel machine learning backend connected to a frontend 3D rendering application. Leveraged Generative AI to enhance workflows while handling the core engineering manually- Enterprise Data Modernization: Led data architecture and engineering efforts to organize interconnected enterprise data from a major corporation for an innovative application within the company. Applied a pragmatic \"right tool for the job\" approach, implementing solutions ranging from simple Excel models to robust, automated data pipelines in Azure/Snowflake- Tech Stack: Azure, Azure DevOps, Snowflake, python, SQL, VertexAI, Node.js, Prompt Flow, Adobe Substance, Blender, git, Excel

Education

• Virginia Tech

  B.S. summa cum laude, Honors Scholar, Environmental Policy & Planning

  2007 — 2010

• Michigan State University

  Master of Science (M.Sc.), Agricultural, Food, and Resource Economics

  2017

• Tecnológico de Monterrey

  Español

  2008 — 2008