Jim Karkanias

Vice President, Data Science and Information Technology

How did you become interested in science?

I was an avid reader as a child, and I was fascinated by how everything worked, particularly the human brain.  I read everything I could, ranging from the back of cereal boxes to books on physics.  As a child, one of my favorite sets of books were the “How and Why” series.  One was dedicated to the brain and I still remember the very last line of the book.  It said something like “…and these three pounds of tissue, which is your brain, is reading and understanding this sentence.” My 8-year old mind was blown.

The passion generated by that experience fueled many years of study in neuroscience. I was very interested in understanding the logic behind how the machinery of life and biological systems work. Popular science fiction talks about advanced nanomachines performing this or that function.  But nanotech is already here, the machinery underlying biology is arguably the most advanced nanotech we have encountered yet.

I started off in primary biological research, and moved into neuroscience and eventually neural networks and software design. I taught myself computer programming as soon as computers were available and ended up working in machine learning long before it was called that; one of the first programs I wrote automated the creation of D&D characters – and using predictive analytics could generate a fitness score for that character. This was in 1979.

I was lured away from academia to the private sector when I joined the pharmaceutical industry.  I soon realized the consultants I was hiring knew about half of what I knew but were making three times as much money.  So, at the beginning of the dot com boom, I left my safe corporate job and helped start a software company where I specialized in advanced analytics.  One of the cooler things we did was image processing and data analytics on physical checks to make pay/no pay decisions. Remember in those days, hardly anyone was thinking about using machines to analyze data in this way. The system learned which features of a check could identify fraud. I cashed out of this space in the late 90s and returned to the pharmaceutical industry but this time to operate a startup group within a corporate environment; that organizational experiment succeeded beyond everyone’s wildest expectations. After that I joined Microsoft to help start a division dedicated to applying advanced information management to life sciences and healthcare. We were so successful that portions of it spun out into their own companies and the primary technology invented there is still bearing fruit in some of the most advanced products being shipped today.  I have the dubious distinction of being named the “grandfather of azure machine learning” at Microsoft. I’m proud of my team’s achievements in the machine learning and information management space; PowerQuery, Azure Data Framework, and of course Azure Machine Learning to name a few.

What was your educational path?

I have a Bachelor’s degree in neurobiology and behavior. I managed to orchestrate an interesting graduate program for myself as a collaboration between Drexel, The University of Pennsylvania and Temple. It was a PhD program that grouped bioengineering, neuroscience and machine learning.  It was certainly one of the favorite times of my life.  I completed the PhD coursework but chose to provide for, and devote time to, my family and then-newborn daughter at the time.

What brought you to CZ Biohub?

The proposition of what Biohub aimed to achieve was very interesting, so much so that they successfully recruited me out of retirement. The more I interacted with the amazing people at the Biohub, the more impressed I became. They really set audacious goals: How do these fundamental biological systems work? How can we repair them? How can we improve them?  Not only were the goals audacious, but they had the means and the will to achieve them. It’s a make-no-apologies approach to changing the world.

This role is an interesting confluence of the skills I have accumulated throughout the years. There is an academic component to my skills, there is the science-oriented work, the neuroscience and biology, then the industrial side of me: operations, pharma, the business of science, and then there is the high-tech side of me – software and machine learning. This role is an intersection of all those things.

When you aren’t working, what do you like to do?

I have a lot of hobbies and my own “lab” at home where I play around with fabrication (3D printing, laser cutting), electronics (arduino mostly but some basic pcb etching and surface mount stuff too), and of course software development.  I’m an avid gamer too. Finally, I’ve practiced martial arts on and off for 30 years. Most recently I started learning Aikido and Iaido, which I appreciate for the peacefulness and discipline at the center of each practice. I try to read about 2 books per month. I play several musical instruments poorly but I am getting better all the time.

Educational Background

  • BS – Rutgers University
    Neurobiology and Behavior
  • MS – Drexel University
    Bioengineering and Biomedical Engineering
  • University of Pennsylvania
    PhD coursework completed in Neuroscience