Kaskada Joins DataStax: Creating the World’s Best Real-time AI Experience Together
2 weeks ago
Blog | 3 years ago | 4 — 6 mins
What if understanding people was the goal?
Meet Charna, a data science lead at Kaskada, and get insights on how to be a more responsible, ethical data scientist (and why the best meals are with good company).
I think for me, my favorite food isn’t really a single dish, but rather the experience of sharing food with other people. I particularly enjoy potlucks where the food comes with a story. I also like to recreate someone's childhood nostalgic food. For get togethers, I'll ask what folks are nostalgic over and then make some version of that.
Right now, I’m nostalgic for my mom's coleslaw, it was cabbage based and always served cold – a classic summer food. She would make it from scratch and use some version of her own blend of vinegar and mayo sauce. She always winged it, so I don't know what the recipe is.
I have so many hobbies, but I'm a human connection, people driven person. And during the pandemic, it's been pretty hard to connect with people. So, I've picked up a lot more of my stay at home hobbies around building furniture and things like that.
Today for fun, I've been doing different types of woodworking, things like cutting boards or built in benches and bookcases. Once we get past this pandemic, I think for fun, I'd like to do some type of bonfire gathering, with people, without distancing.
I would say most people wouldn't know how much time I spend reading about other people's points of view. I had someone once tell me that we have to be careful with the thoughts and ideas that we consume. They believed that exposure to ideas was dangerous. I’m the opposite, without understanding what folks in a different position are hearing, seeing, and doing, I don’t know how to connect with them and discuss ideas. A great example of this is politics and the news. I would much rather read news as facts, then left and right leaning analysis of those facts. This way I can understand the emotional state other folks might be coming from and how to even talk to them about the topic.
I would say two things. One – data science is inherently applied in my mind. So, the best place to start using data science is on a topic that you already know a lot about. You have the human heuristics; from the thousand times you've done something. You think you know this is what works. See if you can try to guide your exploration of the data with hypotheses and then prove and disprove the things that you thought you already knew.
The second thing would be to study things like behavioral science and ethics. The thing that we're trying to do with every model and product we build is change a behavior and there are unethical traps you can fall into. And oftentimes these ways of testing and thinking aren't built into data science classes. There’s a great book by one of my friends, Matt Wallet, called Start at the End on products, behavioral psychology and what kind of ethical checks that we can do when seeking to change behavior. I would recommend giving the book a read or two then try to apply it to what you're doing right now.
My path started in digital signal processing in the defense industry, I chose signal processing because it applies to every field. I spent the first 10 years of my career individually applying these techniques to different projects.
When I left defense, I joined the startup world in Seattle, at a startup trying to provide feedback to writers on how the words we use impact people in real time, as we write. And that was the first time I had focused on a single field using machine learning to build that product. Although you could argue that communication is at the heart of everything that we do.
When I was trying to figure out what I wanted to do next I learned about Kaskada, the platform was going to enable domain experts, data scientists, to get more of their models and features into production, faster. I think the stat is that 80% of our data science models don't get into production. And joining Kaskada was such a compelling next step, making an impact on the entire world of data science by advancing the way we work.
My path has always been in some form of engineering, whether it's electrical engineering, computer engineering, et cetera. I've been doing data science for a long time, under different titles.
But I didn't know when I started school that I was ever going to go to college. And I didn't know when I went to college that I was going to get a PhD. And when I got a PhD, I had no idea whether or not I was going to be in academia or in industry. Each step of the way I made the decision that kept as many options open and had the greatest impact, and data science has the capacity to make a huge impact.
Along the same line of thinking is the ability to unlock the potential of data science and to allow data scientists to spend less time cleaning data and spend more time on considering how to use features to get the biggest positive impact on the people they are impacting.
There's an entire field of study, dedicated to ethics and I believe if data scientists can spend less time on cleaning data and translating models, we can spend more of our time understanding the behaviors we’re trying to change and the problems we're trying to solve.
Written by Max Boyd, Data science lead
2 weeks ago
A month ago
Free and Open Source Software (FOSS) is software whose source code is freely available for anyone to read and modify. Since its origins in the GNU a ...Read More
2 months ago
Dear Readers, this blog post is contributed by our guest writer, Loretta Jones, VP Growth atRead More
2 months ago
This is the second part in our “How data science and machine learning toolsets benefit from Kaskada’s time-centric design” series, illustrating how ...Read More