About Me.

About Me.


  • Leadership View LinkedIn

    Versatility is the center of how I lead. I can take the latest foundation or vision model—or a specialized Hugging Face model, tool, or dataset—understand its real capability, and adapt it to the problem a business actually needs solved. I have worked with and led teams ranging from seven to nearly fifty people, building the ML and AI applications that help organizations learn quickly, move confidently, and accomplish their goals with ease. I can grow and align a department when the work demands it, but I still prefer the focus, ownership, and building velocity of working solo.

  • Applied AI Architecture View LinkedIn

    I turn model capability into dependable products. That can mean evaluating frontier and open-source models, using Roboflow, or reshaping a Hugging Face dataset for a narrow domain. I have built my own applications to label, format, train, and validate data so every model receives the inputs it actually needs. I also enjoy finding new value in existing research data—for example, repurposing Structured3D for computer-vision problems beyond its original use case. Whether I am building retrieval and agent tools or the evaluation, security, and observability needed to ship them, the goal is practical: give people a faster, smarter way to work.

  • Machine Learning Application View GitHub

    I have applied ML to healthcare pricing, document intelligence, recommendation systems, computer vision, legal risk, and autonomous retraining. I enjoy the stubborn middle of the work: testing models against real failure cases, tuning them to the domain, measuring what changed, and staying with the problem until the result is useful—not merely impressive in a demo.

  • Firmware Engineering View Projects

    Firmware development is more than a skill—it’s my creative playground. Whether I’m coding motion algorithms for an XGO Lite robot dog, prototyping smart home tech like a voice-controlled mirror, or deploying object-tracking systems across a Coral TPU, I love pushing the boundaries of what’s possible at the edge. I’ve built virtual assistants using ChatGPT, and even created a networked edge-computing camera that scans physical invoices and transfers digital records across systems. It’s hands-on, systems-level engineering that reminds me why I started building in the first place.

  • Game-Ready Design View Projects

    I am building two connected systems: Return of the Mesh turns residential plans into reviewed IFC and FBX based game-ready homes, while Holoprojector turns real fixtures and furniture into accurate GLBs. Across them I have trained and evaluated models over more than a million vector primitives, including ViT, RT-DETR, VecFormer, RoomFormer, PolyRoom, and Gemini-based geometry reconciliation, then compared image-to-3D output from TRELLIS.2, Rodin, Meshy, and Tripo. The biggest lesson is that visual similarity is only the start: correct dimensions, location, topology, pivots, UVs, collision, and instancing are what make an asset truly game-ready. Blender automation and Unreal MCP let me validate all of that quickly enough to build complete, lively worlds without sacrificing runtime speed.

Certifications

  • Azure Solutions Architect Microsoft 2022

    3 part specialization achievement - azure foundations, azure admin associate, azure solution expert. Learned all tools and resources available within the Azure cloud environment. Using tools daily to orchestrate data or machine learning flows

  • Databricks Expert Architect Databricks 2022

    3 part specialization achievement - Databricks Foundation, Databricks Associate Engineer, Databricks Advanced Data Engineer. Learned all tools and resources available within the Databricks cloud environment. Using tools daily to model machine learning pipelines, create new product pipelines, share and network with team on best practices

  • SnowPro Advanced: Architect Snowflake 2022

    2 part specialization achievement - SnowPro Core, Snowpro Advanced: Architect. Learned all tools and resources available within the Snowflake cloud environment. Using tools daily to sell data on the marketplace, aggregate for reports, or drive a new product pipeline

Specializations

My Person