James Orr
Chief Product & Technical Officer
.png)
"I’m a very intuitive person. If I try to be overly rational and put all the pros and cons in a spreadsheet, I get myself into trouble. But if I follow my gut, I rarely make mistakes."
What I believe
What I've learned
I spent a lot of time on the internet as a teenager in the 1990s. It was all new back then, and there was a gold-rush vibe to it. Every company wanted to have a presence there. And I became that kid who knew how to make websites. That’s how I got my start in tech.
Later I was an early employee with Questica. I just dove in, and in about six months I completed a large piece of what is now the enterprise finance system used by hundreds of governments. All around the world, my code is still running in the background.
I love learning and I love pianos. I play for two hours a day, and I will take my ARCT exam this August. I love the music, but it also helps me think laterally. I’ve worked in computing all my life. This is a different way to use a keyboard.
New technologies tend to grab the world’s imagination and light it on fire. We think it will connect us all and help us build a new utopia. Technology never does that. But it always turns out to be very helpful and useful for very specific things. You have to see through the hype.
What’s next
Artificial intelligence today is like the internet back in the 1990s. There’s lots of hyperbole. But there’s also an incredible amount of human effort going into it, great minds everywhere pouring themselves into building it. There are big new innovations almost every week. Many of them are obsolete by the following week. It’s dizzying.
We plan to become a world-class data organization, setting the standard for data practices in the AI age. AI can act as an agent - a personal trade expert - on behalf of the user that can use specialized knowledge to construct a complex search that extracts an insight from billions of rows of trade data, then turn around and teach the user how and why it did that. This is the sort of expertise we rely on from our research department, but as an agentic AI system it'll scale.
It's not just user-facing innovation, either: we're also using language models on the back end to enrich data in our pipelines. We can identify real business entities in misspelled text blobs, infer product codes and categories from vague descriptions, and many other enhancements that require a semantic, rather than algorithmic approach.It's hard to imagine what our software will look like in a few years, let alone software systems in general, but it's an exciting time to be in this business.