Tyler Alika Gee
I build AI and ML systems from research to production deployment.
timeline
2025 - Present
Working in AI and data roles tied to financial services at Sunwest Bank, focused on production use cases.
2024 - 2025
Data Scientist/AI Engineer at Datafy in Ogden, Utah. Built AI-driven analytics for geolocation and video insights, anomaly detection models, reinforcement learning systems, and MLOps workflows for continuous model improvement.
2018 - Present
Data Scientist/AI Engineer at the University of Utah (part-time/research track), leading and building ML systems for medical diagnostics, data pipelines, and edge deployment on NVIDIA Jetson platforms.
2022
Data Scientist/AI Intern at Micron Technology (Boise). Built visualization tooling, manufacturing analytics dashboards, and ML pipelines for metrology prediction.
2021
Process Engineer Intern at Marathon Petroleum (Salt Lake City), shipping process optimization work that improved refinery operations and throughput economics.
education
B.S. Chemical Engineering, University of Utah (3.3 GPA).
A.S., Weber State University (3.8 GPA).
bio
Tyler Alika Gee is an AI and data scientist with experience across healthcare, semiconductor manufacturing, finance, and edge AI. His work spans computer vision, anomaly detection, reinforcement learning from human feedback, and production MLOps.
current and featured projects
Current builds are sourced from live product sites, GitHub repositories, X profile links, and resume entries (as of April 20, 2026).
active products and ventures
- ValdrSystems - Founder-led web services for high-stakes firms, from audit and strategy through launch.
- SearchForge - Execution-first GEO platform for detecting citation gaps and shipping search visibility fixes.
- Ordalis - AI-powered PDF-to-Excel and document data extraction product with API-driven workflow automation.
- Common Compute - Batch AI compute platform for embeddings, transcription, OCR, rerankers, and image captioning.
- appsprintai - Early-stage project in your GitHub portfolio; public positioning is not yet surfaced on a live site.
- Keko'olani Reunion - Reunion registration and logistics site for family event planning, updates, and attendee coordination.
- Waipio Cultural Experience (repo) - Demo experience site for the Waipio cultural offering.
- Graderhood - Platform for rating HOAs and neighborhoods for the greater good.
- OhanaAI (repo) - Ongoing AI initiative and public hub for your applied AI work.
private and proprietary systems
- Iris (private) - AI for Mac with native app surfaces, assistant workflows, and deep local automation.
- MLX Diffusion (private/research) - Diffusion-model experimentation and image-generation workflows optimized for Apple Silicon and MLX.
open-source and research implementations
- TANGRAM - Robotics scene-understanding and simulation pipeline (resume: project in progress).
- Makakilo (NDA) - Computer vision interface that processes uploaded video and derives tabular behavioral metrics.
- mlx-turboquant - Apple MLX quantization and compression work for efficient local inference.
- DeepGen (repo) - Genealogy research platform integrating retrieval, evidence tracking, and AI-assisted updates.
- VerilogDeepSeekR1 - Text-to-Verilog model adaptation stack with supervised and reinforcement-learning training flows.
- SNN-MLX - Spiking neural network experiments built on Apple MLX.
- Kaleo - Hawaiian speech tooling pipeline for text-to-speech and speech processing.
Need help shipping AI systems for your company? Work with me via ValdrSystems.
highlights
- AI team lead on medical diagnosis collaboration with UCSF using breath-sample data and ML/CNN methods.
- Built custom MLOps pipelines for iterative retraining and deployed ML classification workflows to Linux-based NVIDIA Jetson edge devices.
- Built AI tooling across SQL, pandas, Django, Tableau, and cloud systems (AWS + Azure), with production-oriented data pipelines.
featured talks
- Guest Lecturer - David Eccles School of Business, University of Utah.
publications
- GC-MS and electro-chemical breath-data classification research paper (in progress).