Me
I’m a software engineer with three decades of experience in a variety of industries and variety of roles up to CTO of the online division of a global company. Want to know the details, check out my LinkedIn profile.
Most recently I’ve been working on my own thing. I really like working with numbers and wish data science and machine learning had been subjects when I was in university. I took the one undergrad course offered in AI in which neural nets were treated as mere curiosities.
I started working with AI professionally in 2015 using decision trees for auto-grading bespoke garment measurements using the Weka framework, and built the data science and machine learning practice at Essilor-Luxottica, launching the first machine learning features on their online retail platform and realizing a $2MM lift on $110M revenues for two engineering weeks’ of effort! And this prior to the emergence of LLMs!
As an executive I felt at arm’s length from the technology and wanted to roll up my sleeves and get my hands dirty again, not spend my days on org charts, budgets, and roadmaps. I am at heart an engineer. I love nothing more than building systems and tinkering with technology. The itch needed to be scratched. Hence …
Pyyka
Pyyka is the name of the automated trading system I am developing using machine learning. I chose a subject that was both challenging with high upside potential on success. The core premise is to apply the same underlying technology as LLMs to fintech. To be clear it is not using LLMs for stock picking. It goes one level deeper into the neural net technology (transformer architecture) but quantitative rather than linguistic.
The pyyka name derives from pika, a small animal resembling a short-eared rabbit found in mountainous regions of the West Coast. The pika has a distinctive cry that sounds like a dog’s squeaky toy. The spelling is of course an artifact of finding a domain name that isn’t parked or already taken.
The Blog
The blog communicates my experience of three decades in the industry starting as a software engineer, moving through technical management, and in to executive leadership, up to my current experience building pyyka. It is intended to be anecdotal and conversational. I hope my experiences resonates with others.
My goal is to keep blog posts easily digestible by limiting word count to 1,000-1,500 words per post.