Jason Grey - AI Resume

Email: jason@jason-grey.com
Web: www.jason-grey.com
Phone: (612) 229-6559
Location: Edina, MN, USA

History Link to heading

Not your typical AI researcher…

I have a long history of working with AI in various capacities. My first exposure to neural nets was in 1994, but my real journey began in the early 2000s when I started exploring the intersection of AI, marketing, and customer experience in the advertising industry. Over the years, I have been involved in numerous projects that have pushed the boundaries of what is possible with AI in the areas of computational linguistics, natural language processing, computer vision, and many other forms of machine learning.

While not classically trained in mathematics or AI, I have always been fascinated by various forms of advanced mathematics (linear algebra, calculus, statistics, probability, fractals, chaos theory, dynamical systems, topology, and ai), and have studied them until I felt confident in my ability to apply them to my projects.

I’m familiar with most commonly used tools and models, and enjoy learning (or creating) new ones.

Projects 2007-Today Link to heading

Internet Broadcasting - very deep computation linguistics at scale (20GB/hour news aggregation engine) - including scratch training of numerous models to precisely identify entities, classify topics, and find precise addresses. This project also involved heavy use of early GIS query engines, and an iOS app launched week 1 of the app store’s existance. This project also used very early AWS (when they had 3 services - EC2, S3, and network), the beta version of Apache Nutch, and early versions of Hadoop.
OLSON - head of technology, innovation and UX - recomendation engines, computer vision & augmented reality for clients such as Target, Nike, Bauer Hockey, 5/3 Bank, United Health Group, and many more.
Target Store of the Future Lab
2nd Swing Golf - Club Identification Model - 700k+ high resolution images of clubs, 2M records, and ability to recognize the type of club, brand and model of 10k models. Used in customer experience and back of house operations.
Airborne Athletics - Computer vision and machine learning for basketball training performance analysis and training recommendations.
NVIDIA - neural & hybrid search and recomendation engine for Video on Demand and site search across 100+ websites.
Verizon - built a model to find outdated marketing materials across the company’s vast network of stores.
Utah Avalanche Center - envisioned unified data platform and prediction pipeline for avalance danger models.
Common Crawl - I did a 1 year contract assisting the CTO with the most advanced work, involving their 9+ petabytes of data.
Kagi - semantic search at (small/clean) web scale (and more)

Innovation, Art & Experiments Link to heading

I am named in a patent for inventing a novel method for generating feedback on basketball shot dyanmics and training recomendations.
rooftop solar nature poet
multi-person poetry driven, musician impacted performance art
personal research project on conversational bridging technique
(stealth project) - for anthropologically based socioeconomically informed agent profile generation for use in AI research and testing. (TLDR: make realistically acting fake people which can have structured conversations, arguments, votes, decisions - for use in research and testing of AI systems)

Thanks! Link to heading

Thanks for reading - I hope you find me well suited for connecting the dots and innovating between highly technical subjects/teams and practical applications to suit the needs of businesses and/or humans. I can go reasonably deep with the acedemics and bring wide and creative perspective to complex challenges involving AI.

I’m very interested in using AI for good… need more content here…

If you’d like more traditional information about me and my work, check out the following links: