Expanding Your Horizon: A Podcast on Business and Technology · Horizon Managed Services

What Is Data Science And How Can It Be Used In Business?

·42 min·2 clips

Arian says a model can never be 100% accurate, and businesses still need a human element.

The episode opens with an interview between the host and Arian Oseman of Evolve24. Arian starts with his background. He earned a bachelor's degree in mathematics. He later completed a master's degree in computational sciences with a focus in computational mathematics. He is finishing a PhD in computational mathematics. Much of that work took place at George Mason in Northern Virginia. He says he worked full time while going to school part time. His professional path moved through software development, engineering, database administration, business intelligence, architecture, implementation, and development. He also worked as an individual contributor in data science. More recently, he has taken on management roles while still staying close to technical work. The conversation uses that background to frame data science as something rooted in business needs, not just theory. Arian explains that a model is only useful if it is being monitored after deployment. He notes that predictions can become invalid once real-world conditions change. That leads into a broader discussion of artificial intelligence. He draws a distinction between a data scientist and a machine learning engineer. In his view, the data scientist is more concerned with the mathematics and statistics behind the model. The machine learning engineer is more focused on flow, implementation, and getting the model into a live environment. He expects the future to bring fewer data scientists and more machine learning engineers. He also points to broader use of AWS, Google Cloud, and Databricks as part of that shift. The episode keeps coming back to the idea that business stakeholders need clear translation, not jargon. Toward the end, the conversation uses fraudulent transactions as a concrete example. The point is not just whether a system catches fraud. The point is whether it does so accurately, whether it mislabels real users, and whether the bottlenecks in the process are understood well enough to address them.

As heard by us

A straightforward look at data science that stays tied to business risk and production reality.

The episode keeps its attention on what data science looks like once it has to operate inside a business, not just in theory. Arian Oseman offers a grounded account of getting into the field, then draws a useful line between model-building on paper and the harder work of putting…

Read the full review in PlayNext →

Why you'd press play

Want a clear read on data science and where AI work is headed?

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Expanding Your Horizon: A Podcast on Business and Technology