pcapp
Location: Bellevue, WA (Seattle area)
Remote: Yes, or on-site/hybrid in Eastside/Seattle-area
Willing to relocate: No
Technologies: TypeScript, Python, Rust, C++, Kubernetes, Docker, GCP, SQL, Neo4j, LLM evals, OpenCV, ITK, Keras, Unreal Engine, Houdini
Résumé/CV: https://www.petercappetto.com/resume
GitHub: https://github.com/pcapp
Email: [email protected]
Current projects: Hobgoblin, a hand-written Rust coding agent that speaks the Agent Client Protocol (https://github.com/pcapp/hobgoblin); Scout, a TypeScript job-search agent that scores postings with several LLMs and stores them in Neo4j; a ComfyUI node pack for visualizing CT scans (https://github.com/pcapp/comfyui-dicom-volume); and a three-node Raspberry Pi Kubernetes cluster (https://vitals.petercappetto.com). I'm also taking Hamel Husain and Shreya Shankar's AI Evals course.
Most recently I was an Engineering Manager on Pokémon GO's internal tools at Niantic. Hands-on, I cut tooling traffic by over 80% by adding compression to an unmaintained Kubernetes operator that manages Envoy, and I architected the move of admin tooling off the game server so it could ship outside the game's two-week release cycle.
Before that, five years at PrecisionOS (surgical VR): I built the CT-scan-to-3D-mesh pipeline and the accuracy bench test behind the product's FDA Class II approval, then led the data platform team. Earlier: games and game servers in Unity, Flash, Scala/Akka, C#, and Java.
I have worked with artists much of my career and also enjoy digital content creation in motion design and 3D (currently exploring Cavalry and seeing how Blender compares to Houdini).
Location: Bellevue, WA (Seattle area)
Remote: Yes, or on-site/hybrid in Seattle
Willing to relocate: No
Technologies: TypeScript, Node.js, React, Python, Rust, Go, Java, Scala/Akka, C++, C#, Kubernetes, Docker, GCP, SQL, Neo4j, Claude Code, MCP, Agent Client Protocol, OpenCV, ITK, TensorFlow/Keras
Résumé/CV: https://www.petercappetto.com/resume
Email: [email protected]
Most recently, I was an Engineering Manager for Pokémon GO's internal tools at Niantic. I architected and led the move of admin tooling off the game server into its own Kubernetes/Node.js deployment, so it could ship without waiting for the two-week game release cycle. I also led the build of a CDN-backed asset upload platform that cut infrastructure costs by $15k/month. Hands-on, I cut tooling traffic by over 80% by adding compression to a largely unmaintained Kubernetes operator that manages Envoy deployments.
Before that, I spent five years at PrecisionOS, a surgical VR company. I built the pipeline that turns CT scans into 3D meshes in Unreal and wrote the accuracy bench test that supported the product's FDA Class II approval. I also did R&D on 3D deep learning segmentation, then led the data platform team (TypeScript, React, Zod). Earlier, I built games in Unity and Flash, and game servers in Scala/Akka, C#, and Java.
I'm currently building agent tooling: a coding agent written in Rust that speaks Zed's Agent Client Protocol, and Scout, a TypeScript job-search agent that scores postings with several LLMs and stores them in Neo4j. I also recently built a three-node Raspberry Pi Kubernetes cluster, using agents as tutors and reviewers along the way. Live status: https://vitals.petercappetto.com
Remote: Yes
Willing to relocate: Yes, PNW ideally.
Technologies:
- Languages: Python, C++, TypeScript, Scala, and many others (OTcl, anyone?)
- AI/Data Science: Jupyter, NumPy, Keras, TensorFlow, PyTorch, ITK, OpenCV, SQL, LangChain
- DCC: Unreal, Houdini, After Effects
Résumé/CV: https://www.linkedin.com/in/petercappetto
Website: https://www.petercappetto.com
Email: [email protected]
I have fourteen years of professional experience focused on complex R&D within gaming and medical technologies. I spent much of the last five years working in computer vision. I have a very holistic view of technology within business.
My ideal role would involve combining AI/computer vision and visualization or real-time interaction.